the tower undergraduate research journal volume v, issue ii

Want to break from the crowd?there’s a lab for that.

WWW.GTRI.GATECH.EDU/CAREERS/STUDENTS

JOBS FOR STUDENTS

The Georgia Tech Research Institute

(GTRI) is a highly regarded applied research

and development organization. Founded

in 1934, GTRI is an integral part of the

Georgia Institute of Technology, one of the

world’s top-ranked research universities.

With more than 1,500 scientists, engineers

and other professionals, GTRI helps

solve the most difficult problems facing

government and industry across the nation

and around the globe.

www.gtri.gatech.edu

THETOWERUndergraduate Research Journal

at the Georgia Institute of TechnologyVolume V, Issue II

The Tower is an interdisciplinary research journal forundergraduate students at the Georgia Institute of

Technology. The goals of our publication are to:showcase undergraduate achievements in research;

inspire academic inquiry; and promote Georgia Tech’s commitment to undergraduate research endeavors.

THE EDITORIAL BOARD

Tyler J. KaplanEditor-in-Chief

[email protected]

Mohamad Ali Najia Managing Editor of Submissions & Review

[email protected]

Katie Staples Managing Editor of Production

[email protected]

Michael Bonifacio Business Manager

[email protected]

Harsimran MannAssociate Editor of Submissions & Review

Section EditorsAmrita Banerjee

Christopher Harper

Bradley WillifordWebmaster

STAFF

FACULTY REVIEWERS

Dr. Tibor Beseds EconomicsDr. Wayne Book Mechanical Engineering

Dr. Amy D’Unger History, Technology & SocietyDr. Monica Halka Honors Program

Dr. Melissa Kemp Biomedical EngineeringDr. Narayanan Komerath Aerospace Engineering

Dr. Don Lim Aerospace EngineeringDr. Lakshmi Sankar Aerospace Engineering

FACULTY ADVISORY BOARD

Dr. Rosa I. Arriaga Interactive ComputingDr. Jeffrey A Davis Electrical & Computer Engineering

Dr. Amy D’Unger History, Technology & SocietyDr. Monica Halka Honors Program

Dr. Marlit Hayslett Georgia Tech Research InstituteDr. Pete J. Ludovice Chemical & Biomolecular Engineering

Dr. Milena Mihail College of ComputingDr. Lakshmi Sankar Aerospace Engineering

Dr. Han Zhang College of Business

SPECIAL THANKS

The Tower would not be possible without the assistance of the following people:

Dr. Karen Adams Fellowships OfficeMr. Charlie Bennett Library

Mr. Jon Bodnar LibraryMs. Beth Bryant Office of Development

Dr. Rebecca Burnett Literature, Communication & CultureMr. Michael Chen Former Editor-in-ChiefDr. Steven Girardot Assistant Vice Provost

Dr. Marlit Hayslett Georgia Tech Research Institute (GTRI)Mr. Van Jensen Georgia Tech Alumni Magazine

Dr. Kenneth Knoespel Literature, Communication & CultureDr. Pete Ludovice Chemical & Biomolecular Engineering

Dr. Kathryn Meehan Fellowships OfficeDr. Carole Moore Assistant Vice Provost for Academic Affairs

Ms. Catherine Murray-Rust Vice Provost for Learning Excellence and Dean of Libraries

Dr. Thomas Orlando Chemistry & BiochemistryMs. Susan Parham Head,

Scholarly Communication and Digital CurationMr. Mac Pitts Student Publications

Dr. Christopher Reaves Faculty Advisor,Undergraduate Research Opportunities Program (UROP)

Mr. John Toon Enterprise Innovation InstituteDr. Lisa Yaszak Literature, Communication & Culture

Dr. Han Zhang College of Business

Production TeamChristopher Cassidy

Shawna HagenFitrah Hamid

Victor LeeJung Min Lee

Nabila NazaraliGraham Rhodes

Undergraduate Reviewers

Casey AultmanKevin Daffon

Jaheda KhanamLaura Lanier

Matias LeguizamonAng Li

Michael MerrittShamus Moran

Simisola OludareChristopher Pace

Hifza SakhiKunal Shah

SeungHo ShinHoki Tse

Graduate ReviewersAmulya ChiralaShereka Banton

Pritha BagchiShriradha Sengupta

Chris QuintoJenna Wilson

Research ScientistDaniel C. Jones

The staff of The Tower would also like to thank the Georgia Tech Board of Student Publications, the Undergraduate

Research Opportunities Program, the Georgia Tech Library Information Center, Auxiliary Services, and the Student Gov-

ernment Association for their continued support.

63 | The TOWER | Volume V, Issue II

LETTER FROM THE EXECUTIVE VICE PRESDIENT FOR RESEARCH

Dear Reader,

As Georgia Tech’s Executive Vice President for Research, I provide support and leadership for all research, economic development, and related activities within the Institute. As a member of the Georgia Tech President’s Cabinet, I par-ticipate in strategic planning, goal setting, creation of new research centers, development planning, and outreach to alumni. Each and every one of these duties has allowed me to understand the unique value of undergraduate research.

At Georgia Tech, we believe that taking prudent risks is critical to the successful development and implementation of our research strategy. In fact, many of our greatest risk-takers are students, and when their creativity is combined with the talents of faculty and staff from different disciplines–as well as the expertise from industry partners, the results can be amazing.

One way our students can be seen pushing the limits of innovation is through the student research competitions we hold each year in collaboration with industry. Another way is by participating in important publications like The Tower. These opportunities allow students to showcase, submit and publish their research in a public forum, and to experience - first-hand - the peer-review process. I strongly encourage all undergraduate researchers to take advan-tage of these unique opportunities.

Over the past five years The Tower has grown from a well thought out idea in need of incubation into a well-developed and respected journal that publishes the best and brightest of Georgia Tech’s undergraduates. I have truly enjoyed watching the progress of The Tower in the past few years and look forward with great anticipation to where the new Editor-in-Chief, Mohamad Ali Najia, will take the organization.

I am excited to introduce to you the second issue of the fifth volume of The Tower. The last volume included a number of changes in layout, design and content. News briefs and interviews were heavily expanded, and the physical design of the journal itself was completely re-worked. In this latest issue, the staff of The Tower has chosen to further cultivate the news briefs section, and expand the diversity of submission types. This issue includes an abstract, synopsis and dispatch in addition to the regular full articles, allowing readers to experience the full spectrum of research publica-tions. Each type of manuscript provides a different perspective and point of view of the incredible research that is being done by the undergraduates at Georgia Tech.

I hope you enjoy reading this issue, and that you will join me in congratulating outgoing Editor-in-Chief, Tyler J. Ka-plan and his staff on a fantastic year, and two wonderful journals.

Go Jackets!

Stephen Cross, PhDExecutive Vice President for ResearchGeorgia Institute of Technology


Dear Reader,

It is both with great anticipation and emotion that I write my final “letter from the Editor” for this journal. After five years I have finally graduated from our beloved Institute, and after spending four of those years working with The Tower Undergraduate Research Journal, it is finally time for me to move on. My time at The Tower has been full of surprises and rewards as I have had the opportunity to work with some of the most talented and awe inspiring individuals in the country. Georgia Tech is world renowned as an epicenter for some of the nation’s most intelligent undergraduates, and The Tower tends to attract the brightest of those individuals.

My staff and I began this year with ambitious goals and aspirations – and I truly believe that we have accomplished them all. We successfully constructed a sustainable financial model that allows us to publish at least twice per year. We successfully established an article exchange program that has fostered new relationships with other universities around the country. We successfully doubled our submissions, and began to publish other forms of research publica-tions such as abstracts, dispatches and synopses. We successfully redesigned the journal to incorporate both artistic elements and elements of a professional peer-reviewed journal. None of these goals could have ever been accom-plished without the tireless work of each division of our staff.

I want to take a moment to recognize in particular the members of the Editorial Board. To Michael Bonifacio, our Business Manager, I thank you for stepping up to the plate and taking over the Undergraduate Research Kalei-doscope for me. To Harsimran Mann and Chris Harper, the leaders of our review staff, I thank you for taking my obsessive management of the review process in stride, and for turning around the past two journals in record time. To Katie Staples, our patient, kind and lovely Managing Editor of Production, I want to thank you for meeting every single one of my absurd requests with a smile, and for designing the most beautiful journals this organization has ever produced. Finally to Mohamad Ali Najia, my lieutenant and now our new Editor-in-Chief, I want to thank you most of all for showing the same dedication and determination that each of the past leaders of this organization have demonstrated. Being Editor-in-Chief is not always an easy task, but I have faith that you will pour your heart and soul into The Tower, just as Michael Chen, Chu Yi, Mark Youngblood and Dianne Palladino previously did.

I would also like to take a moment to thank those who have mentored and advised me along my journey as Editor. Dr. Carole Moore (who is featured in this journal), Mr. Mac Pitts, Ms. Marlit Hayslett and Mr. Van Jensen all deserve special recognition. Finally, to the readers, I want to thank you for your loyalty and support. We all do what we do for you, and hope that you enjoy the journals that we work so hard each year to produce. As for my post-graduation plans, I will be staying in Atlanta to work as a lobbyist for the JL Morgan Company, a lobbying and public affairs consulting firm that represents client interests at all levels of government.

Yours in white and gold,

Tyler J. KaplanEditor-in-ChiefThe Tower Undergraduate Research Journal

LETTER FROM THE EDITOR-IN-CHIEF


TA B L E O F CO N T E N T S

6 7

N E W SGeorgia Tech Mathematician Enhances the Predictive Power of a Biodiversity Indexby Hi f za Sakh i

7 1 The Invention Studio: Turning Ideas into Realityby Hi f za Sakh i

6 9 The Entrance into the Stem Cell Era: An Opportunity for Therapeutics, Diagnostics, and Drug Discoveryby Mohamad Al i Na j i a

6 8 The Origin of the Ribosomeby Amri ta Baner j e e

7 2 The Tower Turns Five Years Oldby Hars imran Mann

7 3

I N T E R V I E W S

S P OT L I G H T : J a c o b R . M c G r e wPrevious Author and Accomplished Undergraduate Scholarby Mohamad Al i Na j i a

7 4

F E AT U R E : D r. C a r o l e M o o r e Assistant Vice Provost for Academic Affairs by Tyl e r J . Kap lan

8 5

A R T I C L E SBIOMEDICAL ENGINEERINGSpectromic Analysis of Sphingolipids in Macrophage Cell Line Grown on Indium-Tin Oxide Reveals Activation of Inflammatory Pathwaysby Chr i s t ine Amuzi e

8 9 ECONOMICSAn Analysis of Georgia’s Venture Capital Exodus and Related Legislative Effortsby Michael de la Guardia

9 7 ELEC TRICAL ENGINEERINGEvaluation of a Tongue Input Method for Control of Environmental Appliancesby Daniel Burnham

1 0 5 PHYSICS Effect of Magneto-Hydrodynamics and Radiation Pressure on the Reionization of Dwarf Galaxies by Martin Halicek

1 1 1 PUBLIC POLICYUnited States Citizenship and Immigration Services U Visa Certification Denial by Law Enforcement: An Obstacle for Immigrant Victims of Crimeby Michael Crino

COMPUTER SCIENCEAlgorithms for Vehicle Routing in Warehouse Settingsby Kyl e Dav i s and Dani e l Hul l

8 1

A B S T R AC T

7 5 PUBLIC POLICYMeasuring Firm Perception and Reaction to the Risks of Climate Changeby J J O’Br i en

D I S PATC H


G e o r g i a Te c h M a t h e m a t i c i a n E n h a n c e s t h e Pr e d i c t i v e Po w e r o f a B i o d i v e r s i t y I n d e x

Biodiversity refers to the variation of life forms in all levels of ecological orga-nization; from the biomes comprised of numerous ecosystems that harbor a rich variety species and communities to the entire biosphere of planet Earth. Biodi-versity, though traditionally a topic of interest solely for ecologists and envi-ronmentalists, has recently gained main-stream momentum as information about global warming and conservation efforts flood into the public realm. Numerous organizations such as the World Wildlife Fund (WWF) are spearheading efforts to protect wildlife and preserve this rich bio-logical heritage of Earth. Amidst this demand for curbing the effects of global warming is the central concern about the erosion of biodiversity. As multitudes of organisms succumb to pressures of global warming and become endangered, there is a perpetual need to assess the impact of global warming on specific ecological communities by quan-tifying their biodiversity. Diversity in-dices are used for this endeavor as they mathematically elucidate both the diver-sity and abundance of a species within a given community by incorporating two important parameters: species richness, m, and evenness. Species richness repre-sents the number of different species in-habiting a particular community, whereas evenness, or equitability, delineates the relative abundance of each species in that community.

The Berger-Parker index, d, is a com-monly used biodiversity index, notable for its simplicity in delineating the pro-portion of the most abundant species. In the formula, X(m) represents the number of individuals in the most abundant spe-cies and n represents the total number of individuals sampled. Hence, the index delineates the proportion of the most prevalent species: lower the value of the Berger-Parker index, the more evenly distributed and diverse the community. However, for a community with a grow-ing number of individuals and species, the typical method of calculating the Berger-Parker index directly is no longer accurate, becoming a barrier for mak-ing accurate biodiversity measurements, which is where the important contribu-tions of a Georgia Tech mathematician come into light.

Dr. Huy Huynh, a recent gradu-ate from the School of Mathematics and present faculty member at the University of Notre Dame has developed statistical tools to estimate the Berger-Parker index as the number of individuals and species grow without bound. While collaborating with his advisors, Dr. Christian Houdre and Dr. Liang Peng, he introduced an es-timator for the Berger-Parker index called the multinomial maximum, through which he obtained a limiting distribution of the estimator as the number of indi-viduals and species simultaneously grew without bound. Next, he constructed 95%

confidence intervals for the maximum proportion, which immensely helped to improve the comparison of the Berger-Parker index among communities. Hence, Dr. Huynh’s research paves the way for enhancing the power of biodiversity mea-surements and has sound implications for the future. For instance, Dr. Huynh seeks to ap-ply these estimation methodologies to bacterial colonies, constituting millions of bacteria. In particular, he would like to estimate the diversity of bacteria popula-tions residing on human skin. The find-ings from his future studies have the potential to help doctors and scientists better understand why some people are more prone to skin conditions like ecze-ma and psoriasis while others with simi-lar genetic backgrounds are not, which could be invaluable for better prognosis and diagnosis. Finding the appropriate data for undertaking this endeavor is his current challenge. Nonetheless, describ-ing biodiversity phenomena rigorously through the language of mathematics can help scientists further elucidate the un-derlying biology, and contribute to pos-sible ecological and medical applications.

H i f z a S a k h i

NEWS

“Descibing biodeversity phenomena rigorously through mathematics

can help scientists further elucidate underlying

biology”


T h e O r i g i n o f t h e R i b o s o m eA m r i t a B a n e r j e e

Here at the Georgia Institute of Tech-nology, there is an organization known as The Center for Ribosomal Origins and Evolution (RiboEvo), headed by Dr. Lo-ren Williams, in which graduate students, post-doctual fellows and research scien-tists work to understand the biochemistry of early Earth. One focus of the RiboEvo group is the origin of the ribosome: a pro-posed ancestral version of the ribosome capable of catalyzing peptide bond for-mation more than 3.5 billion years ago. The ribosome is often referred to as the “Ultimate Frozen Accident” because evolution has left the structure of the ri-bosome largely unaltered. When compar-ing the structure of the macromolecule across different domains of life, such as archaea and bacteria, the ribosome can be broken down into two subunits: the Large Ribosomal Subunit (LSU) and the Small Ribosomal Subunit (SSU).The focus of Ri-boEvo is on the LSU because this is the portion of the ribosome that catalyzes the formation of bonds between amino acids in a growing protein. The RNA that sur-rounds the site of peptide bond catalysis (the PTC) in the ribosome is highly con-

served in sequence and even more highly conserved in structure among modern ribosomes, and so is believed to be the oldest part of the ribosome. The Williams Lab has shown that their proposed ances-tral PTC forms a structure very similar to that of the modern ribosome and retains the ability to interact with proteins need-ed for ribosome assembly. While research on the ribosome is a primary focus of RiboEvo, the laborato-ries also conduct experiments on the com-ponents of the ribosome: RNA, proteins, and divalent cations such as Mg2+. One ongoing project is a study of interactions between RNA, believed to be the first biopolymer on Earth, and Iron II (Fe2+). Before the rise of oxygen (O2) from pho-tosynthesis, the Earth’s waters were rich in Iron II, a soluble form of Iron in the absence of oxygen. RiboEvo has shown that the Iron II metal interacts heavily with RNA. As the concentration of O2 in Earth’s atmosphere increased, Iron II would have become less soluble, precipi-tating out of solution as Iron III, to even-tually be replaced by magnesium (Mg2+), the dominant divalent cation in modern

biochemical systems. In a PLoS ONE pa-per published last year, the Williams Lab demonstrated that Iron II can substitute for Mg2+ in RNA folding. A more recent paper just accepted to Nature Chemistry documents a novel function of certain RNAs in the presence of Iron II, the abil-ity to catalyze electron-transfer. The work of RiboEvo will have a big effect on the origin of life community. The lab’s demonstration that Iron II can replace Mg2+ in biochemical systems un-der simulated early earth conditions, and impart novel function to some RNAs, gives further support to the RNA world hypothesis, which states that RNA arose before DNA and proteins. RiboEvo will continue to impact the scientific commu-nity as research continues forward.

“The ribosome is often referred to as the ‘Ultimate Frozen Accident’ because evolution has left the structure of the ribosome largely unaltered”


NEWS

T h e E n t r a n c e i n t o t h e S t e m C e l l E r aA n O p p o r t u n i t y f o r T h e r a p u t i c s , D i a g n o s t i c s , a n d D r u g D i s c o v e r y

M o h a m a d A l i N a j i a

In recent years, stem cell research has emerged as one of the most exciting areas of scientific discovery and medical prom-ise. Human embryonic stem cells capture the imagination because they are immor-tal and have an almost unlimited devel-opmental potential. After many months of growth in culture dishes, these remark-able cells maintain the ability to form cells ranging from muscle to nerve to blood—potentially any cell type that makes up the human body. The proliferative and developmental potential of human em-bryonic stem cells promises an essentially unlimited supply of specific cell types for basic research and transplantation thera-pies for diseases, ranging from heart dis-ease to Parkinson’s disease to leukemia. Stem cells can also be used to study an individual’s disease progression in vitro, opening up opportunities for personal-ized therapeutics and pharmaceuticals. The early concept about how to har-ness stem cells was simplicity itself: har-vest the unformed cells from embryos and inject them into needy recipients. The stem cells would then start rebuild-ing damaged hearts, pushing cancer to re-mission, or healing injured spinal cords. Heart disease, ALS, Parkinson’s disease, and type I diabetes would all be swept away under the tidal wave of the stem cell cure. However, similar to a foreign kidney or heart transplant, injection of foreign cells would likely cause immuno-

logical rejection by the body. Destruction of human embryos for scientific research also poses numerous ethical challenges. In 2006, researchers in Japan lead by Dr. Shinya Yamanaka from Kyoto Univer-sity had an ingenious yet simple scientific breakthrough. Their discovery, which won the Nobel Prize in Physiology or Medicine, was predicated on reprogram-ming mature somatic cells, such as skin fibroblasts, into embryonic-like stem cells by using viruses to add only four genes into the fibroblast’s nucleus. These re-programed cells or “induced pluripotent stem cells,” iPS cells, were shown to dif-ferentiate into nearly every cell type in the body and evade the perils of rejection since the cells were the patient’s own. iPS cells essentially eliminate all ethical chal-

lenges surrounding the derivation of stem cells because no embryos are destroyed. Numerous studies since have greatly advanced our understanding of the biol-ogy that regulates stem cell differentia-tion and the stem cell microenvironment. However, in order to realize the clinical promise of stem cells, our fundamental knowledge of stem cell biology must be translated into suitable applications. Dr. Todd McDevitt, Associate Professor in the Wallace H. Coulter Department of Bio-medical Engineering and Director of the Stem Cell Engineering Center, argues that the nascent field of stem cell engineering will be increasingly necessary to realize the scientific community’s envisioned goals of stem cell based therapeutics, di-agnostics and drug discovery platforms. Engineering conditions for iPS cells to differentiate homogeneously into a specific cell type is currently difficult. One study demonstrated that although global gene expression of iPS cells looks amaz-ingly similar to embryonic stem cells, there are distinct regions in the genetic code of iPS cells that do not get repro-grammed properly. In those regions, iPS cells’ genomes still resembled the tissues from which they came from, suggesting that the cells had not been fully set back to the embryonic stage. Consequently, iPS cell cultures can become contaminated with other cell types, which do not have the same coveted therapeutic potential.

“Embyronic stem cells capture

the imagination because they are

immortal and have an almost unlimited

developmental potential”


NEWS IN REVIEW

Manually sorting these stem cells is time consuming and difficult; using chemical approaches can damage the DNA inside. To address the problem, post-doctoral fellow Dr. Ankur Singh and Dr. McDevitt, in collaboration with Pro-fessor Andres Garcia have demonstrated a tunable process that separates cells ac-cording to the degree to which they ad-here to a substrate inside a tiny microflu-idic device. The adhesion properties of the human iPS cells differ significantly from those of the cells with which they are mixed, al-lowing the potentially-therapeutic cells to be separated to as much as 99 percent purity. The high-throughput separation process, which takes less than 10 min-utes to perform, does not rely on labeling technologies such as antibodies. Because it allows separation of intact cell colonies, it avoids damaging the cells, allowing a cell survival rate greater than 80 percent. The resulting cells retain normal tran-scriptional profiles and differentiation potential. Using inexpensive, disposable “cassettes,” the microfluidic system could be scaled up to filter increased volumes of cells and to allow the feasible possibility of commercialization and manufacture of stem cell therapies for humans. The Georgia Tech researchers applied their new understanding of the adhesive properties of human iPS cells to develop a quick, efficient method for isolating these medically important cells. Their work, published in the journal Nature Methods [1], represents an innovative conversion of basic stem cell biology, biomaterials, and engineering into a strategy with ther-

apeutic potential. During testing, iPS cell cultures were first allowed to attach to the microfluidic device before being subject-ed to the flow of buffer fluid. Cells with a lower adhesive strength detached from the substrate at lower flow rates. By vary-ing the flow rate, the researchers were able to separate specific types of cells with high purity from mixtures in which those cells accounted for only a few percent of the total. Since their discovery, iPS cells have captured the imagination of researchers and clinicians seeking to develop patient-specific therapies. Reprogramming adult tissues to embryonic-like states has count-less prospective applications in regen-erative medicine, drug development, and basic research on stem cells and devel-opmental processes. To this point, more than 2100 research papers on iPS cells

have been published since 2006, indicat-ing a highly active and rapidly develop-ing research field. Purification of iPS cells marks a significant achievement in realiz-ing the promise of stem cells for therapies and scientific advancement. While much remains to be learned and significant challenges remain in iPS cell research, the development of reprogramming tech-niques represents a breakthrough that will ultimately open many new avenues of research and therapy.

References[1] Singh, Ankur et al. (2013). Adhesion strength–based, label-free isolation of hu-man pluripotent stem cells, Nature Meth-ods, 10, 438–444.

“...development of reprogramming techniques represents a breakthrough that will open avenues of research and therapy“


NEWS

T h e I n v e n t i o n S t u d i oTu r n i n g I d e a s i n t o R e a l i t y

The Invention Studio is a state of the art design-build-play facility located on the second floor of the Manufacturing Re-lated Disciplines Complex (MRDC) and is maintained by the Georgia Tech Makers Club. It is one of the largest design spac-es in the United States and the only one staffed solely by undergraduates, volun-teering to maintain a culture of safety and responsibility while still encouraging cre-ativity, knowledge, and freedom. The Invention Studio was conceived in 2009 by ten mechanical engineering Capstone Design students who had prior manufacturing experience. These stu-dent volunteers gathered to run and op-erate the facility and provide training in exchange for 24 hour access. With more funding from generous corporate spon-sors, the space expanded with more tools and is now open for use by students of all academic disciplines at Georgia Tech. The Invention Studio serves the design needs of students enrolled in 25 courses, such as Senior Capstone Design, Sophomore Design, Dynamics, Control Systems and Dynamics, and Mechatronics. Alongside this, the Invention Studio also serves as a springboard for entrepreneurship. For instance, Invention Studio projects have

been featured in the InVenture Prize and Capstone Expo, both of which create busi-nesses from students’ design engineering. Currently, the Invention Studio spans six rooms and is visited by over 500 stu-dents monthly who come from diverse disciplines such as industrial design, mechanical engineering, electrical and computer engineering, as well as bio-medical engineering among others. The studio features high-end CAD/CAM/FEA workstations and design/build/test workbenches to help students design and build their projects. Among the diverse tools offered by the Invention Studio are some of the most advanced prototyping tools available, including an OMAX wa-terjet cutter, three Trotec Speedy 300 la-ser cutters, fifteen 3D printers, three 3D scanners, CNC lathe and mills, a vacuum former, an injection molding machine, as well as full traditional machine and wood equipment. There are many proj-ects currently in the works at the Inven-tion Studio including a Faraday cage for a particle accelerator, combat robot, jet tur-bine wheels, steel string acoustic guitar, experimental model aircrafts, and a 50 kg satellite. “The Invention Studio is an incredible

endeavor unique to Georgia Tech that is redefining engineering education by giv-ing students a chance to ground their classroom knowledge by building physi-cal prototype” says Xo Wang, president of the Makers Club. “Other elite universi-ties contact us all the time wanting to du-plicate our success in engaging students and effecting change in curricula.” The Makers Club has also funded sev-eral projected, which have been built at the Invention Studio. One such project is the DIY electric guitar built by Tyler An-derson. Tyler, who has been playing the guitar for the past nine years, also has experience with woodworking and ro-botics. He read two books about building guitars before he embarked on his journey to design a custom electric guitar back in the Fall of 2012. Likewise, a custom star tracker to be used with DSLR cameras for astronomy photography is another proj-ect funded by the Maker’s Club and built at the Invention Studio. Students of all majors are welcome to come to the Invention Studio and embark on the maker’s journey to turn their ideas into a reality. To learn more, please visit their website at inventionstudio.org.

“‘The Invention Studio is an incredible endeavor unique to Georgia Tech that is redefining engineering education...’”

H i f z a S a k h i


T h e To w e r Tu r n s Fi v e Ye a r s O l d

H a r s i m r a n M a n n

It is hard to believe that in just five years The Tower has developed from a semblance of an idea from a late night discussion into Georgia Tech’s premiere, peer-reviewed undergraduate research journal. In that time we have published six editions of The Tower; hosted eight Undergraduate Research Kaleidoscopes; interviewed 35 professors, students, and a Nobel Laureate; trained nearly 100 un-dergraduate and graduate students to critically review technical literature; and created article exchange programs with universities around the United States. The founding members of The Tower endeavored to create a forum to inspire, promote, and showcase undergraduate re-search at Georgia Tech. Today, The Tower provides state and national opportunities for many Georgia Tech undergraduates passionate about research to publish and present their work, supporting the Insti-tute’s emphasis on student innovation and discovery. In recent years, the number of submis-sions received for publication has nearly doubled, resulting in lower manuscript acceptance rates into our print journal (Figure 1A). However, in Fall 2013 we will start publishing online-exclusive articles in order to increase our publication vol-ume. In recent years, online viewership of The Tower has dramatically increased, more than tripling since publishing Vol-ume I (Figure 1B). In comparison, we only print 3000 copies of each volume a semes-ter. Thus, online publishing offers greater visibility for our authors and universal

accessibility to the journal’s content. The Tower attracts a diverse array of research fields (Figure 1C) providing readers with a didactic and holistic view of Georgia Tech undergraduate research. More importantly, authors benefit from interdisciplinary collaboration not only for its immediate benefits in a research project, but also for opening diverse ca-reer paths. Isaac Newton was quoted say-ing, “If I have seen further than others, it is by standing upon the shoulders of gi-ants.” The Tower, in serving as a forum to host diverse ideas, will continue to cata-lyze interdisciplinary collaboration into further intellectual inquiry. In a similar light, two-thirds of authors published in The Tower pursue graduate school to seek advanced degrees in their field (Figure 1D). Remarkably, 80% of au-thors that attend graduate school receive the National Science Foundation Gradu-ate Research Fellowship, becoming their fields’ next leaders and experts. One such author and Electrical Engi-neering alumnus, Santiago Hassig, completed his master’s degree in Opto- and Nano-electronics with a thesis on metrol-ogy in industrial calibration pro-cesses, focusing particularly on surface response methodologies , data mining, sig-

nal processing, and data visualization. As a previous undergraduate researcher, Hassig cites that his publication in The Tower “made him think about the big pic-ture.” “Knowing how to appeal to a wide audience without diluting the value of the work is something that I first experienced when I submitted to The Tower,” he says. Santiago is currently a Field Program En-gineer at Schlumberger and a scholar in Society of Petroleum Engineers. As we look ahead to the next five years and beyond, the staff of The Tower remains inspired by the success of our au-thors and motivated to delivery only the highest quality content for our readers. Publishing in The Tower never signifies the end to a research project, but rather the beginning of a journey graced with opportunities. We at The Tower are simply happy to be part of that journey.

Figure 1. [A] Acceptance rates of manuscripts per year. [B] Number of unique views of the online version of the journal. [C] Distribution of manuscripts by academic disipline. [D] Proportion of post-college plans of authors


INTERVIEWSSPOTLIGHT

Assistant Vice Provost for Academic Affairs and Chair of the Student Publications Board

Dr. Carole Moore

Ty l e r J . K a p l a n

Although Dr. Carole Moore chose to attend school in Santa Barbara for the surfing, she came out the other end with an excellent education from the Univer-sity of California. As a senior she thought she wanted to teach social studies, so she went on to get a masters and PhD under the direction of C. Warren Hollister. Out of college her first job was at the University of Notre Dame, which was the first year that they integrated women into the University. She remained at Notre Dame for five years before going to work at Georgia Tech (not before founding and coaching the first women’s tennis team at the University of Notre Dame). After being recruited by Miller Templeton, she began her job at Tech at an exciting time under the leadership of Dean Dull. Since that time, her positions at Tech have covered just about every academic aspect of campus. She now works as the Assistant Vice-Provost for Academic Af-fairs. She serves as the director of the Ox-

ford Program—one of Tech’s signature study abroad programs, and is currently the chair of the Board of Publications. Dr. Moore began her work with the Student Publications board in 1995. In this role she serves as an advisor and con-fidant to each of the Editors-in-Chief and their respective advisors of all student-run publications. The Board of Publications is the gov-erning body that oversees the operations of each of the many publications across campus. The Board is also responsible for selecting new Editors and for ensuring that each publication produces the high-est quality work possible. In her position as chair, Dr. Moore works throughout the year with each of the student Editors and faculty advisors as a link to upper level leadership at Tech, and advocates across campus for each of the publications. The Tower, along with many other publications across the country, is be-ginning to face a looming transition to online-only publishing. Dr. Moore con-tends that The Tower serves as an excellent ambassador for Georgia Tech across the country, and would not be nearly as effec-

tive in that vein as a solely online publica-tion. When asked to characterize chang-es in The Tower over the past five years, Dr. Moore describes a fledgling publica-tion that began as a bold idea with little traction. “In the years since its inception, the work of the staff and leadership at The Tower has grown it from an elemen-tary document to a polished, profes-sional work of which the Institute can be proud,” she states. Dr. Moore also offers sage advice to undergraduates considering research at Georgia Tech by saying, “Get out of your comfort zone and seek out opportunities that will position you well for your future endeavors. Nothing serves that purpose quite as well as undergraduate research.” Each of the past Editors-in-Chief of The Tower is sincerely indebted to Dr. Moore for her years of service. She has been an enormous asset to the develop-ment and consequently the successes of The Tower over the past five years.


Previous Author in The Tower andAccomplished Undergraduate Scholar

Jacob R. McGrew

Fostering and maintaining collabo-ration is a core principle in the Georgia Tech research environment. We at The Tower feel the same and were founded on this belief. One of our largest initiatives—predicated on this belief—has been to enter article exchanges with peer institu-tions around the United States. Our very first article exchange was with the Uni-versity of Oregon, featuring undergradu-ate researcher and fifth-year senior dou-ble majoring in Economics and Music and minoring in Mathematics, Jake McGrew. Jake’s article published in The Tower, entitled “United Way of Lane County’s Promise Neighborhoods and the Ben-efits of Reading Readiness,” focused on measuring statistical relationships be-tween defining characteristics of incom-ing kindergartners and their initial lit-eracy scores. He and his faculty mentor, Elizabeth Lohrke, found that a child’s improvement and learning skills could be as important, if not more so, to literacy scores as strict reading readiness at the start of kindergarten. Their work has led

to several public policy recommendations for Oregon’s United Way of Lane County, which are entering the legislative conver-sation. Besides his publication in The Tower, Jake contributed to the 2013 report up-date of “Building a Grad Nation” while working as a research fellow at Civic En-terprises in Washington D.C. the summer of 2012. He is currently working closely with Dr. Bruce Blonigen of the University of Oregon on a paper entitled “Tasks and Trade Policy Preferences.” Jake has won multiple scholarships and awards as an undergraduate economics major, includ-ing the Grace Miller Scholarship and the W.E. Miner Chair Award for Distin-guished Honors Research in Economics. Jake works as an Economics tutor for a variety of students, and is currently em-ployed by the University Teaching and Learning Center as well as the Athletic Department’s Services for Student-Ath-letes. He has been a research assistant to Dr. Bruce Blonigen for two years, while also completing research tasks for other University of Oregon economics faculty upon request. In 2012, Jake worked with United Way of Lane County analyzing the effects of the newly implemented Promise

Neighborhood programs in Springfield, Oregon and Eugene, Oregon. Jake is on track to graduate with hon-ors from the University of Oregon in June 2013. After serving as a Peace Corps vol-unteer in Eastern Europe from September 2013 to December 2015, he plans on go-ing to graduate school on the East Coast to obtain a PhD in Economics with an emphasis in education, health, and labor. With this degree, Jake intends to either become an economics professor or an an-alyst working in a public policy firm. He hopes to focus his research on the effects of early educational systems on domestic and international economies. Being an avid saxophonist, Jake plans on continu-ing to play in local groups throughout graduate school as well as later in life. He cites that his publication in The Tower will help him “immensely in the pursuit of en-tering a prestigious graduate program.” We at The Tower wish Jake the best in pursuit of his career aspirations. Editor-in-Chief Tyler Kaplan would also like to extend a personal congratulations to Jake on his graduation from the University of Oregon, and for his many accomplish-ments throughout his undergraduate ca-reer.

FEATURE

M o h a m a d A l i N a j i a


ABSTRACT

M e a s u r i n g Fi r m Pe r c e p t i o n a n d R e a c t i o n t o t h e R i s k s o f C l i m a t e C h a n g e

J J O ’ B r i e n 1 a n d D a n i e l M a t i s o f f 1, P h D1S c h o o l o f Pu b l i c Po l i c y, I v a n A l l e n C o l l e g e o f L i b e r a l A r t s , G e o r g i a I n s t i t u t e o f Te c h n o l o g y

INTRODUCTION The Carbon Disclosure Project (CDP) is a nonprofit organiza-tion based out of London, England, which partners with over 550 investors with assets of over $71 trillion to encourage major companies to voluntarily disclose greenhouse gas emissions in an annual survey. The CDP organizes the surveyed firms accord-ing to ten different industrial sectors ranging in diversity from the Energy to Financials sectors. Certain questions within the CDP’s annual survey ask firms if they feel at risk of the physical or regulatory effects of climate change. A different set of ques-tions asks firms if they have taken action during the surveyed year to reduce their carbon emissions. Through a comparison of the responses to these two sets of questions, we attempt to answer the larger question: Is a firm’s concern for climate change necessarily an indicator of whether or not that same firm en-gages in methods to reduce their carbon emissions? This study provides insight into the kind of information that we may re-ceive from the CDP and similar voluntary surveys pertaining to greenhouse gas emissions.

METHODS We analyzed data from CDP surveys for the years of 2008, 2009, and 2010. In their surveys, the CDP targets a non-random sample of firms representing the largest and most successful firms across the world. The CDP groups firms according to ten industrial sectors: Consumer Discretionary, Consumer Staples, Energy, Financials, Health Care, Industrials, Information Tech-nology, Materials, Telecommunications, and Utilities. The CDP divides risk into two categories, regulatory risk and physical risk. Regulatory risk is a perceived risk of penalties and other

negative results stemming from regulations and legislation ac-companying climate change. Physical risk is a perceived vulner-ability to the tangible effects of climate change, such as flooding, natural disasters, crop devastation, or other negative effects. For our analysis, we averaged the two risk categories for each sector to produce the “Risk” category that is presented in our results. Firms that have implemented carbon reduction plans are la-beled as “Action Taken”. Individual firm responses to questions were aggregated according to industrial sector. This process was repeated for each sector and for the questions pertaining to whether or not firms had taken action to reduce their carbon emissions.

RESULTS We found that, while some sectors exhibit a high percentage of concerned firms with a low percentage of firms participat-ing in emissions reduction plans, some sectors were exactly the opposite (Figure 1A). While certain sectors, such as Consumer Discretionary and Materials have fairly consistent percentages for “Concern” and “Action Taken,” other sectors, such as Energy and Health Care are much less consistent. Additionally, most sectors seem to indicate similar trends for all three years. That is, if they exhibit much more “Concern” than “Action Taken” (or vice-versa) in one year, they are likely to exhibit the same pattern in the other two years. We attribute this to differences across sectors in terms of feasibility and affordability of partici-pating in an emissions reduction plan. Given that the CDP has a non-random sample of respondents, we did not test for statisti-cal significance in these results.


Figure 1B and Figure 1C illustrate two cases of sectors behaving differently than would be expected given their perceived risk. In Figure 1B, the Energy sector perceives a degree of risk greater than the amount of action taken in all three years. Conversely, Figure 1C shows how the Health Care sector behaves oppositely – more firms take action than those who perceive risk. We attri-bute these behaviors to the differences in the cost of mitigating greenhouse gas emissions among the various sectors. For exam-ple, it is a much larger and more expensive endeavor to replace or update greenhouse gas emitting facilities associated with the Energy sector than it would be to take action in the Health Care sector.

CONCLUSIONS Based on the results from these three years of the CDP, not all firms which exhibit concern for the risks of climate change

also engage in measures to reduce their own carbon emissions. Likely, this is due to the relatively high cost of reducing carbon emissions in certain industrial sectors. In sectors where the costs are not as high, we examine some firms engaging in carbon re-duction activity despite the fact that they do not perceive them-selves to be at risk. Therefore, we conclude that firm concern for climate change is not necessarily an indicator of whether or not that same firm engages in methods to reduce their carbon emissions. This information sheds light on the larger question pertaining to what kinds of information we can gather from the CDP and similar voluntary greenhouse gas reporting initiatives. Future research initiatives may examine other questions in the CDP or other similar initiatives to further our understanding of what we can learn about participating firms and their reactions to carbon reduction and the threat of global climate change.

Figure 1. (A) Concern versus Action Taken, plotted as a percentage of firms responding to questions of interest. (B) Some Energy Sector firms that feel at risk do not take action against climate change. (C) Some Health Care Sector firms that do not feel at risk still take action against climate change.


©2012 Southern Company

National Carbon Capture Center in Wilsonville, Alabama.

Idea Power.™

We’re not just creating power. We’re creating jobs. Southern

Company is committed to generating a new kind of energy–

an energy of science, technology and research–that will

provide affordable and sustainable electricity for generations

to come. We are looking for the next generation of researchers

and engineers. To see how Southern Company can be part of

your future, go to SouthernCompany.com.

IBEW Local 613 Electricians Do Everything from Wiring Electrical Outlets

to Programmable Logic ControlsYour electrical system is the lifeblood of your business. If it fails, your phones

don’t ring, computers won’t work. Your BUSINESS STOPS.

Don’t take chances with your electrical system. Demand the best trained

Electrical Professionals available from an IBEW Local Union 613 Electrical

Contractor. IBEW Local 613 members complete a rigorous 5 year classroom

and on the job training program. They know how to get your electrical

system back up and running as quickly as possible.

Power Up. Call IBEW Local 613 for the name of Qualified Union Electrical

Contractors near you.

International Brotherhood of Electrical Workers Local 613

501 Pulliam Street, SW Suite 250 Atlanta, Georgia 30312

(404) 523 8107www.ibew613.org

Max Mount Jr

President

Gene R. O'Kelley

Business Manager

DISPATCH

ABSTRACTVehicle routing algorithms can be used to route packages through large networks, analyze and help improve traffic patterns, and be used to improve the efficiency of shipping warehouses. Working with several shipping companies we have tailored our algorithms to achieve the best performance under the latter application. Topology of the warehouse, distribution of the jobs, ordering of the jobs, and the vehicle routes influence the efficiency of a warehouse. Our goals were to determine which aspects of the problem influenced performance the most. After running simula-tions changing one aspect of our solution strategy at a time we determined which parts of the problem were most important. The purpose of this is to lay a solid foundation of algorithms and determine which ones should be further explored and improved upon. The results of the simulations pointed toward the most important factors being the topology and method for routing vehicles. How the jobs were distributed among robots, and how they were ordered once distributed seemed to have little impact on the overall performance. Most of our initial efforts were not focused on the topology. These results could lead to more research focused solely on how to construct topologies that allow for rapid delivery of packages.

Ky l e D a v i s 1, D a n i e l H u l l 1 , 2 , Te t a l i Pr a s a d 3 , 4 , 5 , P h D, a n d H e n r i k C h r i s t e n s e n 6, P h D 1 S c h o o l o f C o m p u t e r S c i e n c e , C o l l e g e o f C o m p u t i n g, G e o r g i a I n s t i t u t e o f Te c h n o l o g y2 H . M i l t o n S t e w a r t S c h o o l o f I n d u s t r i a l & S y s t e m s E n g i n e e r i n g, C o l l e g e o f E n g i n e e r i n g, G e o r g i a I n s t i t u t e o f Te c h n o l o g y3 A l g o r i t h m s & R a n d o m n e s s C e n t e r a n d T h i n k - t a n k4 S c h o o l o f C o m p u t e r S c i e n c e , C o l l e g e o f C o m p u t i n g, G e o r g i a I n s t i t u t e o f Te c h n o l o g y5 S c h o o l o f M a t h e m a t i c s , C o l l e g e o f S c i e n c e s , G e o r g i a I n s t i t u t e o f Te c h n o l o g y6 R o b o t i c s a n d I n t e l l i g e n t M a c h i n e s C e n t e r, S c h o o l o f I n t e r a c t i v e C o m p u t i n g, C o l l e g e o f C o m p u t i n g, G e o r g i a I n s t i t u t e o f Te c h n o l o g y

A l g o r i t h m s f o r Ve h i c l e R o u t i n g i n Wa r e h o u s e S e t t i n g s

INTRODUCTION Vehicle routing has been a challenging problem since the advent of self-driving vehicles and robots. This problem is in-teresting due to its wide range of applications. Vehicle routing algorithms can be used to route packages through large net-works, analyze and help improve traffic patterns, and be used to improve the efficiency of shipping warehouses. Working with several shipping companies we have tailored our algorithms to achieve the best performance under the last application. Every-thing from topology of the warehouse, distribution of the jobs, ordering of the jobs, and the vehicle routes influence the effi-ciency a warehouse can achieve. Given a topology and a list of packages, the objective is to move the packages from their source to destination while mini-mizing the total duration. A topology is a 2-D grid containing

sources, destinations, and vehicles. Vehicles are not allowed to collide with other vehicles, sources or destinations. The list of packages is given in a specific order and describes the source each package is originally located at, and the destination it must travel to. Usually there is a partial ordering that must be main-tained when delivering packages, the packages must arrive in the order given as input to each destination. The problem has been relaxed to ignore the ordering restriction as well as allow-ing the vehicles to have infinite acceleration. This means that vehicles can accelerate to their maximum (unit) velocity and de-celerate to zero velocity instantaneously.

METHODS All simulations were coded in Java making heavy use of the MASON Framework [3]. This framework has many built in


tools for visualizing different types of simulations. We made use of built in tools to render the simulations in order to visualize the interactions of robots with the different topologies. Each to-pology was constructed by hand, and iterated on to improve the performance without changing the overall structure of the to-pology. The most difficult part of getting the simulations to work was coding it in such a way that it was possible to swap out algo-rithms easily and quickly. There was special care taken to keep the code general enough that we can develop and swap in more algorithms easily, but not so general that it would be impossible to code in a reasonable amount of time. A functional interface and factory pattern were used to accomplish this. The functional interfaces have no state and just serve mutate the problem based on the algorithm the interface represents. Then a factory is build that can return different implementations of these interfaces de-pending on which method you call. This allowed for a healthy amount of generality in our simulations.

RESULTS There were many aspects to this problem. Our initial re-search was focused on determining which parts of the problem had the greatest impact on overall efficiency. This will pave the way to more focused research related to the most important parts of vehicle routing, with the goal of implementing these so-lutions and improving the performance of warehouses.

Topology When designing a topology there is a tradeoff between space and speed. With more space, vehicles have more room and avoiding collisions becomes easier, allowing for the faster transportation of packages; however, space is expensive and

Figure 1. Proposed topologies where sources are black and destinations are red. Vehicles are blue circles, indicating it is currently carrying a package, or red circles, indicating that it is not carrying a package

Figure 2. The performance of each topology, provided the num-ber of robots and number of transportation commands were con-stant for each topological simulation

Figure 3. Performance of first-fit and random job allocation to robots


cannot be sacrificed without significant increases in perfor-mance. We tested four different topologies that were built with approximately the same space efficiency but the sources and destinations rearranged. The number of vehicles was held con-stant throughout our testing. Figure 1 represents the different topologies with the vehicles in the process of running their re-spective routes. When holding other aspects of the problem constant there was a wide variance in performance when changing topologies (Figure 2). Performance was measured by the duration of time (unit time step) the robot would take to reach its destination. Topology 3 (Figure 1.3) had the best performance due to its cen-trally located destination bins. Topology 4 (Figure 1.4) had the worst performance due to the polarity of the destination bins, causing robots to travel longer distances, on average.

Job Distribution A job does not have a specific vehicle it must be assigned to. Distributing the jobs to the vehicles is another part of the prob-lem that must be addressed. If each job’s completion time can be estimated, then to minimize the total time, jobs should be evenly distributed among all robots. This minimizes any given robots total completion time. This can be modeled as a 1-D bin pack-ing problem. Each robot is a bin with a maximum capacity. Pack the bins with jobs using their estimated completion time, adding more robots as necessary. In order to conform to our original problem some additional work is required. A binary search is performed to find the mini-mum, maximum capacity that does not require more than the predetermined number of robots. The packing using this max-imum capacity is then given as the job distribution. A simple first-fit solution was used as the bin packing algorithm. It was

compared against randomly assigning jobs to robots without at-tempting to achieve an even distribution (Figure 3).

Job Ordering Once the jobs are fixed to a robot, the order in which each robot completes its jobs can be altered. Intuitively having all of the robots work on an order in the same part of the warehouse at the same time will causes more congestion than if the robots’ current jobs are spread out evenly through the warehouse for a given point in time. The algorithm we use to evenly distribute the jobs through the warehouse uses a simulated annealing ap-proach (Tsitsiklis et al. 1998):

1. Cache Paths: The first step is to compute and store the paths between every source and destination, taking into account the stationary obstacles of the topology. Let Pi,j(t) be the point on the path between source Si and destination Dj at time t. We will be using these values as estimates since we cannot take into ac-count robots colliding yet, estimates are fine for our purposes.

2. Randomly pick a robot to optimize: At random decide on a robot. For every pair of jobs assigned to the robot swap them, and if they increase the fitness function keep the change with a probability that increases to 1 over time.

3. Compute interesting points: We will define interesting points to be the 3 points location at the start of the time interval, the middle of the time interval, and the end of the time interval. These points in time must be recomputed each time the fitness function updates the arrangement.

4. Fitness Function: The fitness function can be computed by

Figure 4. Performance of stimulated annealing and random job ordering approaches.

Figure 5. Performance of vehicle routing algorithms, using breadth first search and an optimal non-obstacle solution


going through every pair of jobs Ja, Jb with sources and destina-tions Sa, Da, Sb, Db, and summing over every interesting point in time relative to the start of the jobs, t.

∑ distance(Pa, a(t), Pb, b(t))2

Maximizing this fitness function will evenly distribute the paths.

5. Repeat and cool: Repeat starting at step 2 and cool down the annealing until a maximum is found.

We compared this simulated annealing approach to randomly ordering the jobs within robots and the performance was nearly equivalent (Figure 4).

Vehicle Routes To find vehicle routes a simple algorithm that performs a breadth first search (BFS) at every time step is used (Rivest et al. 1990). For each robot, determine where it needs to move to-wards next, perform a BFS using the robot as a starting point. Move along the path if one exists, otherwise do nothing. This very simple algorithm treats robots as stationary objects for each time step, so if a robot is in an aisle blocked off by two other ro-bots it will not move until the other robots are out of the way. We compared this algorithm to an optimal solution where robots were not considered obstacles (Figure 5). Intuitively, the opti-mal solution where robots are not considered obstacles would be expected to perform better. Despite the BFS algorithm having slightly worse performance, it is still within the range of vari-ability of the optimal solution. A BFS algorithm could still be a feasible routing vehicle routing mechanism.

CONCLUSIONS After running many experiments two parts of the problem that were explored had a significant impact on the number of steps to deliver all packages. The most significant factor was the topology. Topologies 2 and 4 required many more steps than topologies 1 and 2. The construction of topologies 2 and 4 in-cluded two separate central delivery locations. This appeared to slow everything down compared to the topologies with one central delivery location. The different strategies of routing vehicles also led to sig-nificant variance in the number of steps required. When colli-sions between vehicles were ignored the optimal routing was achieved for the fixed distribution and ordering of jobs. Using BFS and treating vehicles as obstacles resulted in an average number of steps that were more than 1 standard deviation away from the optimal. Future research can further investigate vehicle routing mechanisms to increase performance. Distributing and ordering the jobs did not significantly im-pact the number of steps required. The solution strategies pro-posed for each part of the problem did not achieve results better than a randomized strategy. This could mean that our strategies are flawed, or these parts of the problem will not be a large con-tributing factor to the overall performance.

References1. John Tsitsiklis, Dimitris Bertsimas. Simulated annealing. Statistical Science, 8(1):10-15, 1998.2. Ronald L. Rivest, Thomas H. Cormen, Charles E. Leiserson. Introduction to Algorithms. 1990.3. http://cs.gmu.edu/~eclab/projects/mason/


ARTICLES

ABSTRACTStudies of inflammation on the molecular level employ bioanalytical assays, including matrix-assisted laser desorption/ionization (MALDI). MALDI technique is used in mass spectrometry (MS) for the analysis of large organic molecules and biopolymers. Current techniques in-volve the manual transfer of cells in culture for spotting on stainless steel, gold, or conductive indium-tin oxide (ITO) glass slides for ion-ization. In recent studies, phagocytic cells demonstrate activation of the MAPK pathway, expression of TNF-α and IL-1, and generation of ROS in response to particulate silica, all indicative of oxidative stress and silicosis. Similarly, lipidomic responses, such as toll-like recep-tor (TLR) activation of RAW 264.7 cells by the lipopolysaccharide Kdo2-lipid A (KLA) mimic aspects of inflammatory bacterial infection. This study aims to streamline cell culture and ionization techniques by plating the RAW264.7 mouse macrophage-like cell line directly on MALDI ITO slides. The inflammatory response due to culture on the ITO slide was compared to KLA-induced macrophage cells grown in the absence of slides. Sphingolipid species abundant in inflamed cells are the C16 subspecies of sphingosine (Sph) and ceramide 1-phos-phate (Cer1P). Though a 1.2-fold difference in Cer1P between KLA-induced positive control cells and PBS-treated cells was observed, the reduction in inflammation was not significant (p = 0.0913). The difference in inflammation as assessed through production of the precur-sor molecule Sph was not statistically significant between KLA-induced and PBS-treated cells (p = 0.4677). The combinations of these two findings provide evidence that necessitates the use of anti-inflammatory MALDI slides for direct cell culture for MALDI-TOF MS.

C h r i s t i n e A m u z i e 1, a n d A l f r e d H . M e r r i l l 2, P h D1 K r i e g e r S c h o o l o f A r t s a n d S c i e n c e s , J o h n s H o p k i n s U n i v e r s i t y2 S c h o o l o f B i o l o g y, C o l l e g e o f S c i e n c e s , G e o r g i a I n s t i t u t e o f Te c h n o l o g y

S p e c t r o m i c A n a l y s i s o f S p h i n g o l i p i d s i n M a c r o p h a g e C e l l L i n e G r o w n o n I n d i u m -T i n O x i d e R e v e a l s A c t i v a t i o n o f I n f l a m m a t o r y Pa t h w a y s

INTRODUCTION Matrix-assisted laser desorption/ionization (MALDI) is a bioanalytical tool that makes many improvements upon tradi-tional mass spectrometry (MS) analysis. The technique employs a crystallizing agent, or matrix, in the absorption and ioniza-tion of the surface of the analyte under UV laser. By doubling the ion flight path, the MALDI time of flight mass spectrometer (MALDI-TOF) increases spectral resolution, allowing for accu-rate compound identification. Additionally, the soft ionization

technique reduces fragmentation of the analyte, further improv-ing the accuracy of precursor masses in compound. Studies of inflammation on the molecular level employ bioanalytical as-says, including matrix-assisted laser desorption/ionization (MALDI). Such properties make MALDI-TOF preferable to MS in the analysis of complex molecules, such as glycoproteins and sphingolipids. Further, it can be used in vitro for the quantita-tion of disease states for cancer and arthritis through known bio-markers. Often, these biomarkers are lipids—the end result of


inflammatory pathway activation. Inflammation is one of the first ways the primary immune system responds to infection in all organs, including the lungs. Silicosis, the inhalation of fine respirable dust and the result-ing inflammation and fibrosis in lungs, is a primary response in pulmonary tissue and a fatal occupational hazard in construc-tion. On the cellular level, inflammation is a highly ordered pro-cess regulated by the MAPK/ERK cell-signaling pathway, lipid mediators, and the cytokines TNF-alpha, IL-1, and IL-6 (Xing, 1994). While much has been done to characterize the inflamma-tory pathways associated with silicosis from crystalline silicon dioxide dust (Fubini, 2003), the question of activation by non-particulate silica is yet to be answered. Further, the regulation of lipid production in this disease has yet to be elucidated. This study aims to identify the activity of lipid messengers present in cells grown on non-particulate silica and to determine the oc-currence of inflammation based on the species present. One can hypothesize that the inability of macrophages to phagocytose the non-particulate silica limits inflammation. The lack of a lip-id-mediated response from macrophages grown on conductive MALDI imaging slides would allow for the use of direct plating and imaging for the detection of inflammatory diseases as di-verse as silicosis, atherosclerosis and cancer. MATERIALS AND METHODS

Materials Immortalized mouse macrophage-like RAW 264.7 cells were obtained (ATCC). The Kdo2-lipid A (KLA) standard (Avanti Po-lar Lipids) was acquired. Indium-tin oxide (ITO) slides for cell culture and MALDI-TOF were provided by The Georgia Insti-tute of Technology Bioanalytical Mass Spectrometry Facility. Spectral and visual data were collected with Bruker Daltonics Autoflex Smartbeam-II MALDI-TOF Mass Spectrometer.

Cell Culture LipidMAPS passage procedure PP0000000100 was used

for RAW 264.7 cells. Procedure was modified with the addition of an ITO slide to each of two 150 mm plates. After incubating overnight at 37 deg C, 89% humidity, and 5% CO2, these cells were observed for normal proliferation and minimal apoptosis, which would be indicative of activation of inflammatory path-ways. LipidMAPS protocol PS0000001401 was used to obtain a sonicated KLA working solution. After reaching maximum con-fluence (5e6 cells/plate), cells grown with and without the ITO slides were again split and plated in 10 mL PBS or 10 mL Kdo2-lipid A (KLA). Cells plated with KLA acted as a positive control for the inflammatory response exhibited with PBS on ITO slides. After 10 minutes, the slides were bathed in deionized water for 15 seconds. Plated cells were at 4 deg C for the 15 minutes prior to MALDI-TOF.

MALDI-TOF Protocol for the negative ion mode MALDI-TOF lipid analy-sis called for the use of a sinapinic acid matrix (Siuzdak, 2006). A 2,5-dihydroxy benzoic acid (DHB) MALDI matrix was prefer-entially selected for improved sensitivity and higher tolerance to salts and detergents such as PBS. Cells were taken from dishes with the following conditions and spotted onto a separate ITO slide in 0.5 uL aliquots of sample and 1.0 uL aliquots DHB ma-trix: KLA + ITO slide and PBS + ITO slide. These were compared to cells induced with KLA or PBS and grown without ITO slides (Sims, 2010). Analysis was performed in negative ion mode at 355 nm. Ionization and fragmentation conditions were opti-mized for sphingolipid standards.

LipidMAPS Database The online LipidMAPS Structure Database (LMSD) text/ontology-based search (Dennis, 2012) was used to identify lipid species present on the grounds of lipid category and mass.

Statistical Analysis An unpaired two-tailed Student’s t-test was used in analysis spectra. The intensity (a.u.) of Sph (σ = 152.8) and Cer1P (σ =

Table 1. Identification of sphingolipid species present using LMSD. The online LipidMAPS Structure Database (LMSD) text/ontology-based search was used to identify lipid species present on the grounds of lipid category and mass. Shown are lipid species shared between RAW 264.7 cells exposed to Kdo2-lipid A (KLA) and phosphate buffered solution (PBS).

Solution Variant m/z LM_ID Common NameKLA 1 505.8063 LMSP02050012 Ceramide-1-Phosphate

2 663.9876 LMSP02010013 N-(pentacosanoyl)-sphing-4-eninePBS 1 505.8257 LMSP02050012 Ceramide-1-Phosphate

2 663.9961 LMSP02010013 N-(pentacosanoyl)-sphing-4-enine


1992.5) expression in KLA-induced cells was compared to the PBS-treated cells grown on ITO slides.

RESULTS Preliminary observations of spectra revealed more unifor-mity in KLA-induced cells, whose spectrum tended to be much lower with greater intensity differences (Figure 1). Following MALDI-TOF MS, lipids were identified using m/z values in the LMSD. Peaks corresponding to the MALDI matrix DHB were re-moved from analysis. Spectra for the KLA-induced RAW 264.7 cells exhibiting species indicative of inflammation after 24 hours were compared to the species observed in cells with PBS on ITO slides (Table 1). The most abundant species included Cer1P (m/z = 505.8063) and the ceramide N-acylsphingosine (Sph) (m/z = 663.9876). Similar production of Cer1P (p = 0.0913) and Sph (p = 0.4677) was observed in cells with PBS on ITO slides compared to KLA-induced cells of the positive control (Figure 2).

DISCUSSION Inflammation, measured by the production of sphingolipid Cer1P, was similar in the cells grown on ITO slides with PBS and KLA-induced inflamed cells (p = 0.0913). The same was true for inflammation measured by production of Sph (p = 0.4677) in the cells grown on ITO slides with PBS and KLA-induced in-flamed cells. For this reason the intensity of Cer1P in PBS-treat-

ed cells was still indicative of inflammation (Sims, 2010). The abundance of Sph in RAW 264.7 cells grown on ITO slides with PBS suggests inflammation occurred, possibly because of the isotonic phosphate buffered saline solution rather than because of a phagocytic response to the silica of ITO slides. Lipid spe-cies in the PBS cells included Cer1P (m/z = 505.82567), Sph (m/z = 663.9961), and neutral glycosphingolipids (m/z = 781.9387, 1148.7358). These results are consistent with the inflammatory oxidative stress predicted by previous experiments (Fubini, 2003).

FUTURE WORK Data in this work was collected from a minimal sample size (n = 3); therefore, the findings of this paper can be supported with more replicates via MALDI MS. Findings would be more powerful with the addition of cells grown in Dulbecco’s Modi-fied Eagle Medium (DMEM) with low glucose. These non-in-duced cells would allow for a better comparison to the positive control, KLA. An inclusion of cytokine production (e.g. IL-1, IL-6, TNF-alpha) could be used to reaffirm the correlation between inflammation and lipid production. Additional replicates will be necessary to support these findings; however, the need for anti-inflammatory MALDI slides in direct culture for MALDI-TOF MS is now evident. The inflammatory response observed in the RAW 264.7 cell

Figure 1. MS spectra for RAW 264.7 cells in the presence of KLA or PBS on ITO slides. Preliminary observations of spectra revealed more uniformity in KLA-induced cells, whose spectrum tended to be much lower with greater intensity differences (A). Cells with PBS had less uniformity and greater ion pattern differences (B).


line allows for study of a number of other disease and animal models. With a clear correlation between atherosclerosis and inflammation previously established (Ross, 1999), the lipid-me-diated response of plaque formation could be explored. Specifi-cally, recent studies have shown the role of glyocosphingolipid accumulation and production in hypertrophy and atherosclero-sis.

Acknowledgements The author would like to acknowledge the mentorship of Mr. Samuel Kelly, Mrs. Elaine Wang, Dr. Yanfeng Chen and Dr. Al-fred Merrill.

References1. Dennis, EA, et al. “A mouse macrophage lipidome.” J Biol Chem. 285. (2010): 39976-85.2. Dennis, EA, et al. “Text/Ontology-Based Search.” Lipid Me-tabolites and Pathways Strategy: Lipidomics Gateway (2012): LipidMAPS Structure Database (LMSD). 3. Fubini, B, et al. “Reactive oxygen species (ROS) and reactive nitrogen species (RNS) generation by silica in inflammation and fibrosis.” Free Radical Biol & Medicine 2003; 34(12): 1507-16.4. Ross, R. “Atherosclerosis -- An Inflammatory Disease.” New England Journal of Medicine, Vol. 340, No. 2. (14 January 1999), pp. 115-126, doi:10.1056/NEJM1999011434002075. Sims, K, et al. “Kdo2-Lipid A, a TLR4-specific agonist, in-duces de novo sphingolipid biosynthesis in RAW 264.7 macro-phages, which is essential for induction of autophagy.” J Biol Chem. 2010; 285(49): 38568-79.6. Siuzdak, G. (2006). The Expanding Role of Mass Spectrome-try in Biotechnology, 2nd Ed., pp 99-107. MCC Press, San Diego, CA.7. Z Xing, M Jordana, H Kirpalani, K E Driscoll, T J Schall, and J Gauldie. “Cytokine expression by neutrophils and macro-phages in vivo: endotoxin induces tumor necrosis factor-alpha, macrophage inflammatory protein-2, interleukin-1 beta, and in-terleukin-6 but not RANTES or transforming growth factor-beta 1 mRNA expression in acute lung inflammation.” Am. J. Respir. Cell Mol. Biol. February 1994 vol. 10 no. 2 148-153.

Figure 2. Comparison of lipid expression in RAW 264.7 cells in the presence of KLA or PBS grown on ITO slides. Abun-dance of lipid species shared between KLA-induced and PBS cells, ceramide-1-phosphate (Cer1P) and N-acylsphingosine (Sph) shown. Expression of Cer1P was 1.2-fold higher in KLA-induced cells (p = 0.0913) than in cells treated with PBS. Little change (p = 0.4677) was seen between PBS and KLA in expres-sion of sphingosine. The expression of sphingosine was 8.3% less in the PBS-treated cells than the KLA-induced cells.


ARTICLES

A n A n a l y s i s o f G e o r g i a’s Ve n t u r e C a p i t a l E xo d u s a n d R e l a t e d Le g i s l a t i v e E f f o r t sM i c h a e l d e l a G u a r d i a 1 a n d D e n e H . S h e h e a n e 2, M B A1 S c h o o l o f E c o n o m i c s , I v a n A l l e n C o l l e g e o f L i b e r a l A r t s , G e o r g i a I n s t i t u t e o f Te c h n o l o g y 2 D e p a r t m e n t o f G o v e r n m e n t a n d C o m m u n i t y R e l a t i o n s , G e o r g i a I n s t i t u t e o f Te c h n o l o g y

INTRODUCTION Venture capital is a form of investment, where capital is in-vested into early stage, high-growth and high-risk companies, usually, start-ups or high-tech firms. Investor capital is pooled into a venture capital fund, which is then invested into a com-pany in return for equity. Venture capital is a major force for job creation and economic growth (IHS Global Insight, 2009). In fact, in 2008 venture-backed U.S. companies employed more than 12 million people and generated close to $3 trillion in revenue (IHS Global Insight, 2009). Public companies with headquarters in Georgia that were once venture-backed account for 571,749 U.S. jobs and one U.S. job was created for every $16,615 of venture capital dollars invested in Georgia (Connell Dowling, Baker & Slone, 2011). However, venture capital investment has been rap-idly decreasing. For example, from 2008 to 2012 venture capital investment decreased by 47% in Georgia and in 2010 only 15% of investments came from in state (PricewaterhouseCoopers 2013; Connell Dowling, Baker, & Slone, 2011). Added to this, large high tech companies have departed the state, including 1) Movaz Networks, which left in 2006 and had received $202 mil-lion of investment while in the state, 2) Viewlocity, which left in 2002 and had received $85 million of investment and 3) Wave&

Optics, which left in 2005 and received $71 million of invest-ment. These companies alone represent over $350 million of lost investment to the state. Such a large exodus of venture-backed companies demonstrates an inability for Georgia to retain high-growth companies. In the spring of 2013, the Georgia General Assembly passed HB 318, which established the Invest Georgia Fund. Invest Georgia aims to increase venture capital investment in Georgia by partnering with private funds. Currently, there exists little public information that presents a descriptive and quantitative picture of venture capital investment in Georgia. If legislative ef-forts to strengthen in state investment are to be pursued, a quan-titative analysis of current venture capital investment is neces-sary for legislators to understand the problems currently facing Georgia’s venture capital market and its inability to retain high-growth corporations. In this paper we attempt to consolidate and gather existing data on venture capital investment in Geor-gia in order to quantitatively describe the strength of its venture capital market. The strength of Georgia’s venture capital market was measured using a number of criteria including gross dollars invested, investment relative to the state’s GDP, percentage of in-vestment coming from in-state and ability to retain corporations

ABSTRACTThis paper attempts to analyze the effectiveness and necessity of current legislative efforts to bolster venture capital investment in Georgia, specifically the inception of the Invest Georgia Fund. The Invest Georgia Fund, which passed as HB 318 in the 2013 legisla-tive session, is an attempt to help the state retain many of the high-tech start-up corporations that have been leaving the state. Whether a substantial number of these corporations have been leaving the state has been a subject of debate. We quantify the number of high-growth corporations leaving Georgia and consolidate and gather existing data on venture capital investment in the state, so the neces-sity of some legislative action like the Invest Georgia Fund can be addressed. A case study is made of the medical technology indus-try to test for state industry size and state industry investment correlations. From the data gathered, a final policy proposal is suggested.


receiving investment. A formal attempt to quantify moderate to substantial venture capital collecting firms leaving the state was also utilized to evaluate and measure the strength of Geor-gia’s venture capital market. Opponents of state-backed venture capital programs have argued investment is directly correlated to the existing size of industries. Thus, the focus should be to make industries more competitive rather than opting to increase investment and selecting companies to receive funds, so called “winners” and “losers” (Harper et al. 2013). This paper sets out to analyze the validity of this point using the medical technol-ogy industry across each state as a case study, where correla-tion between the two variables, industry size and investment, was measured. The medical technology industry was chosen because 1) large amounts of data describing the industry and medical technology investment size are readily available, 2) the medical technology industry is one of the highest venture capital collecting industries and 3) the medical technology industry at-tracts 22 percent of venture capital investment in Georgia (Con-nell Dowling, Baker & Slone, 2011). From the data gathered, we argue against the point made by opponents of state-backed ven-ture capital funds. Finally, this paper assesses the effect current legislation would have on addressing the previously mentioned weaknesses.

METHODS

Data Aggregation In order to measure the strength of Georgia’s venture capital market, we compared two datasets: 1) venture capital entering the state and 2) venture capital leaving the state in the form of corporate acquisitions, mergers and high venture capital col-

lecting firms that relocated outside the state. The first dataset, gross venture capital investment entering the state, primarily came from PWC’s MoneyTree Report (PricewaterhouseCoopers, 2013). Data that came from this study on gross investment enter-ing the state provided the base for total venture capital invest-ment in Georgia between 2000 and 2007. In an attempt to analyze and quantify investment leaving the state a second dataset was aggregated from public sources, including the Atlanta Business Chronicle, 1998-2012 and the Atlanta Business Wire, 2000-2012. This dataset represents companies that left Georgia between 2000 and 2012 and had received a minimum investment of $2.5 million.

Correlation Analysis In order to measure the correlation of industry size and in-vestment, a case study of the medical technology industry (MTI) was conducted. Linear regression models were developed, where MTI sales by state were correlated to investments. In or-der to make sure the correlation was measured as accurately as possible, three more linear regression models were developed based on choosing slightly adjusted datasets. In the second mod-el all states with MTI industries of a size less than $300 million in sales or investment less than $2.5 million were not considered. This was done to compensate for outliers, where industry or in-vestment is virtually non-existent.

RESULTS

Company Relocation Between 2000 and 2007, 51 companies left the state repre-senting over $1 billion of lost investment. These exodus compa-

Figure 1. Places of company relocation Figure 2. Places of company relocation as a percent of lost in-vestment to the state of Georgia


nies represent 18% of venture capital, private equity and angel backed capital investment in Georgia for the years 2000-2007. A close analysis of the data points out some key issues facing Georgia’s venture capital market. Figure 1 shows a breakdown of where companies relocated. A more telling picture is shown in Figure 2, which breaks down place of relocation as a percent-age of lost investment to the state. Overwhelmingly, companies representing the most investment lost to the state went overseas. This is not a completely accurate picture because of the roughly $450 million that left to go overseas a little less than half was rep-resented by Movaz Networks’ relocation to Germany. California attracted the most companies. Between the years 2000 and 2007 we recorded 16 companies that left Georgia for California rep-resenting over $215 million of lost investment. Florida attracted the second largest number of companies, acquiring 6 companies totaling $80 million. Finally, New Jersey acquired 5 companies totaling $113 million. To determine what influence the location of investors had, a graph was constructed cataloging what percentage of compa-nies relocated to a state that had at least one previous investor (Figure 3). Of the 48 companies where full information on in-vestors was available, 20% relocated to an investor state. This demonstrates at least a low level of correlation between investor location and place of company relocation. Interestingly, the big-gest influence, determining place of relocation within the U.S., was the presence of a start-up ecosystem or cluster. According to Breznitz and Taylor (2008), Silicon Valley, Metro and Great-er-New York and Boston all contain large and strong high-tech clusters. This factor had much more bearing on place of reloca-tion than the investor state. Silicon Valley, Metro and Greater-

New York and Boston by far attracted the most companies.

MTI Industry Case Study Figure 4 clearly shows that there is very poor correlation in the size of a state’s MTI industry and the amount of venture capital investment the MTI industry attracts in each given state. In fact, the average fluctuation, based on the difference between the estimated amount of investment rendered from the best linear regression model and the actual amount of investment, was over $70 million. To give an idea of how large the average fluctuation is, the average amount invested per state was $74 million, so the average fluctuation was just as big as the average itself. The linear model yields a correlation coefficient value of r2 = 0.337, which means there is very poor correlation between industry size and industry investment in this model. Therefore, the amount of investment a state’s MTI industry attracts cannot be accurately estimated by the size of that state’s MTI industry. These findings support the argument venture capital invest-ment is not a function of industry size, but rather a function of other more important variables. Such findings contradict con-ventional wisdom, but further analysis makes sense of the data. This paper proposes venture capital investment is a function of the abundance of high-growth start-ups in a region, not the cur-rent industry size. Areas that have high-density clusters of high growth start-ups will attract the most investment (Braunerhjelm & Feldman, 2006). This can be seen in regions such as Silicon Valley, Boston and Colorado in the case of the cloud computing industry (Saxenian, 1994). In summary, based on the linear regression models show-ing correlation to industry size and industry investment by state this paper argues little weight need be given to the size of a giv-en industry in a state when estimating venture capital invest-ment. The linear regression model show there is little correlation between the size of a states MTI industry and the amount of ven-ture capital collected in that industry. This can be expanded to apply to all industries attracting venture capital. This data also contradicts the points made by many that focusing on making an industry more competitive alone will encourage more ven-ture capital investment. The results from this case study would suggest a focus must be shifted towards encouraging start-up clusters in a state.

DISCUSSION Since 2011, a purposeful effort has been made by state legis-lators in Georgia to introduce legislation aimed at encouraging more in-state venture capital investment. The first piece of legis-lation aiming to attract more venture capital investment in Geor-gia called for the establishment of a Certified Capital Companies

Figure 3. Influence of investor location on place of company re-location


(CAPCO). However, because of CAPCO’s expensive and ineffec-tive results in other states, Georgia legislators started consider-ing alternative models (Barkley, Markley & Rubin, 2001; Rubin, 2011). The Invest Georgia Fund, passed later by the Georgia General Assembly, provides an independent eight-person board of advisors to pick a third party fiduciary to select venture capi-tal funds. The data collected confirms that Invest Georgia would at least partially address the weaknesses in Georgia’s venture capital market discussed in this paper, and that the Invest Geor-gia Fund is a necessary step towards strengthening that market. Many have argued Invest Georgia Fund legislation over-steps the government’s role in the private sector. It is important to note that selection of venture capital funds is several arm lengths away from the state government avoiding cronyism. The presence of a third party fiduciary, which decides on the funds to invest in, is based on a similar program in Maryland. Added

to this, unlike in a CAPCO model, the state bears no part of the risk and is entitled to its original principal, as well as 80% of the profits. This model more accurately follows the standard for most venture capital investor deals. Representative Dudgeon, who helped develop the Invest Georgia Fund and who is one of the Georgia House’s most conservative voices, makes a tell-ing point about state programs aimed at attracting more venture capital - states are competitors and if Georgia does not adopt similar practices when it comes to incentivizing investment, Georgia will fail to compete (M. Dudgeon, personal communi-cation, April 20, 2013). In fact all state governments neighboring Georgia, except Alabama, have already instituted and funded similar programs in their own states. These programs have been luring high-growth companies to leave the state of Georgia. An example of this is Advanced Catheter Therapies, formerly Atlanta Catheter Therapies. The company, which recently an-

Figure 4. Industry and investment correlation with a minimum investment of $2.5 million


nounced a new US Technology patent, left the state and relo-cated to Chattanooga in 2011 because it could not find adequate funding in Georgia. It was able to raise $3 million from Tennes-see based investors (Advanced Catheter Therapies, 2011). One of the investment requirements was relocation to Tennessee. At-lanta has a very competitive medical research environment with the presence of two of the top ranked research universities in the nation, but lack of access to venture capital is causing more and more firms to leave the state (Albers et al. 2012). Invest Georgia carries with it three major functions that will help strengthen the venture capital market in Georgia: 1) it injects more capital into the market allowing “deserving firms” more access to seed money, 2) after firms receive investments from the invest Geor-gia fund they are required to stay in the state for a given period of time; the longer a company stays in a state the more likely it is to never leave, because it develops strong roots in the region, and 3) Invest Georgia provides mentorship to companies that receive investment. These last two functions address the biggest problem facing the venture capital market in Georgia: invest-ment/company retention (see Results). Other policies must shift to addressing company retention, as well. When companies representing 20% of investment for a given period leave the state in just 7 years, a state must shift its legislative priorities. Georgia is losing the majority of its com-panies to technology clusters (see Results), but Georgia itself lacks a well-developed technology cluster. Another way to de-scribe a technology cluster is as an ecosystem for start-ups and high-tech firms (Breznitz & Taylor). A start-up ecosystem is de-fined by mentorship, well-developed cross-company social net-works and sharing of ideas and resources. A study conducted by Breznitz and Taylor (2008) speaks to the most important facet of a well-developed start-up ecosystem - collaborative culture and social networks. Breznitz and Taylor declare no amount of fac-tors or financing would help create a Georgia technology cluster and keep venture capital-collecting firms in-state without social structure. The data gathered in this paper supports this asser-tion. In order to retain start-ups these companies must be locally embedded into the Atlanta/Georgia technology industry social network. However, even when including investors and legal advisors, the overwhelming majority of Atlanta start-ups and high-tech firms shared no connection. This is in stark contrast with other cities collecting large amounts of venture capital and producing and retaining significant numbers of start-ups, such as Silicon Valley, Boston and New York. Currently, there is no social structure in place to help local Georgia start-ups grow, meaning they will not become embedded in the local business community and they will be forced to look out of state for men-

torship, development of ideas and financing. This is the major issue policies must address. Over $1 billion of lost investment is not a number that can be ignored. Many of the exodus companies have gone on to grow much bigger and produce substantial tax dollars for the states they relocated to. Many high-tech, high paying jobs left the state along with these companies. Policymakers must improve the social network of the technology industry in Georgia, by promoting top Geor-gia execs to be personally involved with start-ups. One way of achieving this is encouraging large Tech firms in the state to in-vest in local start-ups. Crafting legislation, which provides some form of tax-breaks to Georgia technology firms that invest in lo-cal start-ups, would be an obvious way to embed start-ups in the local business community making them much more likely to stay. This would address many of the key issues facing Georgia start-ups who feel they are on their own and migrate to tech-nology clusters outside of the state, searching for mentorship, financing and cross-pollination of ideas. Other policies encour-aging cross-company social networks must also be pursued. An alternative solution providing mentorship and cross-company social networks is developing a more locally based venture capi-tal market. As mentioned before, only 15% of venture capital comes from in state. Pursuing programs like invest Georgia help encourage more in-state venture capital investment.

CONCLUSIONS While in gross terms Georgia receives a fair amount of ven-ture capital investment, an alarming percentage of companies are leaving the state. Further, Georgia has many alluring facets to offer start-ups and high tech firms, but the state is underper-forming in these industries. All this speaks to an unhealthy ven-ture capital market. Venture capital firms are a key part of any start-up ecosystem. They provide financing, mentorship and an incentive to stay near. If Georgia wishes to continue to grow and become one of the most economically strong states in the na-tion it must address the problems facing its high tech companies and its local start-ups. This starts with strengthening the weak-nesses in its venture capital market. There is not just one way to do this. This paper suggests several options, but policymakers must continue to put focus on this issue. Georgia policymakers must find a way to retain its start-ups and high tech companies. This can only be achieved by creating better “ecosystems” for start-ups in the state. These ecosystems need to provide men-torship to companies, cross-company social networks, sources for financing and sharing of ideas and resources. Programs like Invest Georgia will encourage more in state capital investment. Though it will not fix all of Georgia’s company retention prob-


lems, it is certainly a good place to start.

References1. 2013-2014 Regular Session - HB 318, Georgia Tourism De-velopment Act; revise certain definitions; provisions. (2013). Georgia General Assembly. Retrieved April 2, 2013 from http://www.legis.ga.gov2. Advanced Catheter Therapies. (2011). New Name An-nounced for Atlanta Catheter Therapies [Press Release]. Re-trieved from http://www.marketwire.com3. Albers, J.; Loudermilk, B.; Byrd, C.; Dudgeon, M.; Sims, B.; Alford, D.; Aschenbroich, A. (2012). Innovate Georgia 2025. The Georgia Science and Technology Strategic Plan 4. Atlanta Business Chronicle. (1998-2012). Book of Lists. 5. Atlanta Business Wire. (2000-2012). Atlanta Business Wire6. Barkley, D.; Markley, D; Rubin, J. (2001). Certified Capital Companies (CAPCOs): Strengths and Shortcomings of the Lat-est Wave in State-Assisted Venture Capital Programs. Economic Development Quarterly, 15, 350-366.7. Braunerhjelm, P.; Feldman, M. P. (2006). Cluster Genesis: Technology-Based Industrial8. Breznitz, D.; Taylor, M. (2008). The Communal Roots of En-trepreneurial-Technological Growth? Social Fragmentation and the Economic Stagnation of Atlanta’s IT Cluster. Georgia Insti-tute of Technology 9. Bureau of Economic Analysis. (2000-2007). Gross Domestic Product by State: Georgia. U.S. Department of Commerce: Bu-reau of Economic Analysis 10. Bureau of Economic Analysis. (2011). Gross Domestic Prod-uct by State: Georgia. U.S. Department of Commerce: Bureau of Economic Analysis 11. Casper, S. (2007). How do technology clusters emerge and become sustainable? Social network formation and inter-firm

mobility within the San Diego biotechnology cluster. Research Policy, 36, 438-455. 12. Connell Dowling, J., Baker, E., Slone, K. (2011). Venture Capital and Entrepreneurship in Georgia. National Venture Capital Association 13. Grantham, R. (2012, December 30). Change in Georgia Pen-sion Law Not Raising Capital Like Intended. Atlanta Journal Constitution. Retrieved from http://www.ajc.com 14. Harper, C. (2013). Blueberries and Venture Capital. Peach Pundit 15. IHS Global Insight (2009). Venture Impact: The economic Importance of Venture-Backed Companies to the US Economy. National Venture Capital Association 16. The Lewin Group (2010). State Economic Impact of the Medical Technology Industry. Advanced Medical Technology Association 17. PricewaterhouseCoopers. (2013). MoneyTree Report. Price-waterhouseCoopers & National Venture Capital Association 18. Rubin, J. (2009, April 3). Testimony on the Washington DC CAPCO Program. Testimony presented at US Capitol Building in Washington D.C.19. Salzer, J. (2011, December 7). CAPCO Investment Law Gets Another Shot. Atlanta Journal Constitution. Retrieved from http://www.ajc.com20. Saxenian, A. (1994). Regional Advantage: Culture and Com-petition in Silicon Valley and Route 128. Cambridge, MA: Har-vard University Press. 21. Skelly, K. (2010, April 16) Where Venture Capital’s Money Came from in 2009. The Wall Street Journal. Retrieved from http://blogs.wsj.com22. Tharpe, W. (2012). CAPCO: A Bad Investment for Georgia. Georgia Budget and Policy Institute (GBPI)


ARTICLES

E v a l u t i o n o f a To n g u e I n p u t M e t h o d f o r C o n t r o l o f E n v i r o n m e n t a l A p p l i a n c e s D a n i e l B u r n h a m 1 a n d M a y s a m G h o v a n l o o 1, P h D1 S c h o o l o f E l e c t r i c a l a n d C o m p u t e r E n g i n e e r i n g, C o l l e g e o f E n g i n e e r i n g, G e o r g i a I n s t i t u t e o f Te c h n o l o g y

INTRODUCTION Interaction with common appliances in the home, office, and outdoor environments is part of our daily lives; however, the interfaces to these appliances may not be suited for populations with severe disabilities, such as Spinal Cord Injury (SCI). The inclusion of assistive technologies (ATs) in everyday living aims to improve Quality of Living (QoL) of those with severe disabili-ties by decreasing their dependency on others for daily activi-ties (Craig, Tran, McIsaac, & Boord, 2005). Our research purpose was to investigate the usability of a new AT by emulating envi-ronmental control scenarios. Technologies that provide remote control and automation features to home appliances are readily available to make these devices more accessible in daily living and more convenient to control (Tao, Zhang, & Wang, 2008). Home entertainment devic-es such as televisions and disc players have been equipped with remote controls for several decades to facilitate manual control without the need to directly interact with the device. Likewise, home-automation systems that control lights, door locks, and garage doors are available on the consumer market and have

been shown to decrease dependency in elderly populations (Cole & Tran, 2002). Control of the appliances located in the home environment using assistive devices is an area of continued interest. Sev-eral human-computer interaction (HCI) methods have been explored, which use the remaining capabilities of persons with disabilities including muscle contractions, brain signals, eye movements, speech recognition, and tongue control (Barea, Bo-quete, Mazo, & Lopez, 2002; Barreto, Scargle, & Adjouadi, 2000; Karmali, Polak, & Kostov, 2000). Some have already been inte-grated with home automation systems (Hawley et al., 2007; Tao et al., 2008). Tongue input is a promising assistive input method as it pro-vides a high level of dexterity and a rapid learning curve (Huo & Ghovanloo, 2010). In addition, tongue input also allows a cer-tain level of privacy and safety due to being hidden inside and protected by the mouth. An AT called the Tongue Drive Sys-tem (TDS) has been developed to take advantage of the benefits of the tongue, providing a wireless, wearable HCI (Xueliang, Wang, & Ghovanloo, 2008). TDS uses magnetic sensors located

ABSTRACTInterfaces to environmental appliances are not well suited for populations with severe disabilities such as spinal cord injury (SCI). A system called the Tongue Drive System (TDS) has been developed to allow control by movement of the tongue, providing a wireless, wearable Human Computer Interface (HCI). The TDS system has been integrated with an Android smartphone for controlling of appliances in the users’ envi-ronments through infrared (IrDroid) and the X10 (X10 Commander) home automation protocols. Both touch-screen and tongue input methods were compared on their effectiveness in the control of a television and lamps. The results of the experiment show that the TDS provides a usable input for the control of common household devices, as each of the subjects was able to complete all of the required tasks. Touch screen performance was 2.96 and 2.05 times faster than TDS input in X10 Commander and IrDroid application, respectively. X10 control had an error rate of 0.27 errors per action and no errors for touch input. The error rate of the IrDroid application was 0.30 and 0.17 errors per action for touch and TDS, respectively. The results show that the TDS can be used in conjunction with a remote controller to carry out actions using environmental devices.


in a wearable headset to track the movements of a magnet lo-cated on the tongue. These movements are then translated into different inputs depending on the location of the tongue in the oral space. By using a smartphone running the Android oper-ating system as an intermediary communication point between the home appliances and the AT input, the control of these appli-ances can be given to a person with SCI or similar disabilities. X10 is an industry standard home automation protocol that uses home power lines as a signal carrier (“X10.com”). The X10 protocol is one of the oldest and most widespread home auto-mation protocols. Many X10 device modules are available, in-cluding appliance controllers, light switches and dimmers. The basic command set of X10 includes: on, off, and dim values. X10 modules generally act as a power switch to the respective ap-pliance that is connected to the module and receives commands from an installed controller. Computer-based controllers are available to allow a server running to control the appliances re-motely and provide a sequence of commands without requiring

direct input. Several applications for Android exist, which leverage the use of a computer server connected to the X10 controller. The “X10 Commander” developed by Melloware® (Media, Pennsyl-vania), which includes an Android application and a compan-ion server runs on a personal computer connected to the X10 power line network (“X10 Commander”). X10 commander soft-ware suite provides a wireless remote control of the appliances equipped with X10 adaptor. Another control method is Infrared. Infrared communication is a line-of-sight remote technology and is nearly ubiquitous in televisions, disc players, audio receivers, and other electronics. Although manufacturers follow different encoding methods, in-frared protocols work similarly by modulating a beam of infra-red light aimed at a device. Dongles that provide infrared transmission are available for mobile devices. One of such dongles is the IrDroid™, which converts the audio signal from the output of an Android device into infrared signal (“IrDroid”). The infrared dongle is accom-panied by software of the same name, which converts the re-mote control code data bits into infrared pulses. The combina-tion of the software and the transmitter allows the smartphone to control infrared devices. The TDS has previously been evaluated as a replacement

Figure 1. TDS tongue positions for various commands.

Figure 2. Setup of experiment showing TDS headset, Android phone and controlled appliances.


mouse pointer on both a personal computer and Apple iOS de-vice (Kim, Huo, Minocha, Holbrook, Laumann, & Ghovanloo 2012). Its use as an electric-powered wheelchair controller has also been previously demonstrated (Huo & Ghovanloo, 2010; Kim et al., 2012). However, the TDS had not been evaluated for its functionality in control of other household appliances. Therefore, the main objective of this work is to evaluate its func-tionality in controlling common appliances in a simulated home environment and analyze its effectiveness using commercially available software and standard protocols. The results indicate how the TDS hardware and software interact with commercially available home automation solutions, and also what improve-ments should be made to the user interface to increase the acces-sibility of devices in the home environment via TDS.

MATERIALS AND MEHTODS

The Toungue Drive System To evaluate the TDS performance in environmental control, six able-bodied subjects were recruited between the ages of 18 to 65 years old to perform several tasks related to controlling home appliances. Finger and tongue input methods were compared to control lights and a television. The finger input utilized the capacitive touchscreen of the Android smartphone. The tongue input used the TDS headset to track a magnet attached near the tip of the tongue to emulate a mouse cursor behavior on the An-droid smartphone. Figure 1 demonstrates the TDS tongue posi-tions and the corresponding commands. During the experiment, subjects controlled two types of commonly used household objects; a television and two lamps. They performed actions on each device including switching the lamps on and off, turning the television on, changing the vol-ume and then turning it off. The IrDroid application and its associated IrDroid infrared dongle were used to control the tele-vision through universal IR codes. X10 Commander was used to control the lamps through the X10 protocol, which were con-nected through an X10 appliance switch to the X10 power line network. Data was collected by manually recording the duration of each task by a stopwatch. To ensure the order of the tasks did not affect the durations, the actions were randomized, while keeping the on and off tasks at the beginning and end respec-tively. The tasks lists were given to each person randomly to eliminate any bias that might be set by the order of the given tasks. In each session, participants performed each using both touch screen and TDS.

Experimental Setup The experiment took place in a room with a table on which a 32” LCD television (Samsung LN32C350D1DXZA) was locat-ed ~1 m in front of the subject, as shown in Figure 2. An An-droid phone (Samsung SPH-D700) running Android 2.3 (Gin-gerbread) was used in the experiment with the TDS Bluetooth mouse pointer software installed. A stand was used to hold the phone to keep it stationary and slightly angled toward the sub-ject during the experiment. The IR dongle was connected to the Android phone, and was pointed directly at the television. Two lamps were also located on the table on each side of the subject. These lamps were connected to one power strip through the X10 CM11A module. The computer server was placed out of sight and connected to the CM11A module. The experimenter sat to the side of the subject so that the operation of the Android tasks could be seen and times recorded.

Procedure The subjects had the magnet glued to their tongue and the external TDS headset was fitted following the procedure in (Yousefi, Huo, Veledar & Ghovanloo, 2011). Subjects then cali-brated the system using the LabView interface to compensate for electromagnetic interference (EMI). Subjects then trained the TDS through a LabView interface in Huo et al 2008, which re-quired the subject to first place their tongue in each of the posi-tions 10 times for accurate placement values. Once satisfactory training was confirmed by observing the command clusters in the virtual feature space, derived via principal component anal-ysis (PCA), subjects were given ~3 minutes to use the TDS to navigate through the Android OS in order to become familiar with it, but were not shown the two Android applications in-volved in the experiment until they performed the main tasks. The experiment included performing eight task lists consist-ing of six actions each. Four of the task lists were performed with the TDS and the other four used touch input. Task lists were chosen randomly. Each task was completed four times us-ing each input method. The first two task lists were used for subject observation; therefore, the subjects were given no infor-mation about how to perform the tasks. These two sets were re-moved from the quantitative results and used to provide quali-tative data based on the errors recorded as the subject learned to perform the tasks. The subjects were given a questionnaire in order to evaluate both the TDS system interaction and the intuitiveness of the Android applications on a Likert scale rang-ing from 1-5 with one being in least agreement and five being in most agreement.


RESULTS All subjects completed all action lists successfully. One sub-ject was removed from the data set due to poor calibration of the TDS headset which resulted in data that was outside of the upper fence, defined as 1.5 times greater than the inner quartile. The first two action sets were considered practice and only used qualitatively to judge intuition about the application interface. The results of the study were averaged across each subject and then and averaged across all subjects. Each action was given an average for the subject and then compared across subjects. Overall the touch performance was 2.96 and 2.05 times faster

than TDS input in X10 Commander and IrDroid applications respectively in terms of time for completion. Touchscreen in-put had a standard deviation of 15.9% of the average for X10 Commander and 12.9% for IrDroid. TDS input standard devia-tion was 60.9% of the average for X10 Commander and 54.1% of the average for the IrDroid. The respective averages for these standard deviations to be compared to are presented in figures 5-8. The lowest subject time for IrDroid was 37.5 seconds with TDS and 24.3 seconds with touch input. For X10 commander, the lowest subject time was 41.6 seconds using TDS and 16.4 seconds using touch input. The overall averages for the action

Figure 3. Average Action times across all users for each action in the IrDroid application action set. Error bars indicate standard deviation.

Figure 4. Average Action times across all users for each action in the X10 commander application action set. Error bars indi-cate standard deviation.

Figure 5. Totals of average action times for each user for X10 Commander using TDS.

Figure 6. Totals of average action times for each user for IrDroid using TDS.


times across all subjects using the IrDroid and X10 Commander applications are displayed in figures 3 and 4 respectively. Fig-ures 5 through 8 demonstrate the individual average action times as a part of the total time taken for each subject. The error occurrence rate, which measures the number of wrong actions performed during the experiment, is presented in figures 9 and 10. The error rate was calculated by taking the number of errors that occurred in the sets and dividing it by the total actions in those sets. The X10 commander application had an error rate of 0.27 errors per action and no errors for touch in-put. The TDS had a total of eight errors in the X10 Commander application. The error rate of the IrDroid application was 0.30 er-rors per action for touch and 0.17 errors per action for TDS. Five errors occurred in the TDS input of the IrDroid with two errors in the IrDroid being system errors. For the touch input, nine errors occurred with six being user errors and the other three be-ing system errors. The system errors in the IrDroid application were due to the unresponsiveness of the television to the power

command. Selected responses to the questionnaire are shown in figure 11. The questionnaire responses indicated that users agreed that the applications were intuitive, though only half of the users agreed they had experience using the Android device. All sub-jects were familiar with smartphones. One user had familiarity with the applications before beginning the experiment. Half of the subjects responded that the TDS was accurate while the oth-ers were mixed between neutral and one disagreed. Four of the users stated they were satisfied with the TDS performance while one was neutral and one disagreed.

DISCUSSION The results of the experiment indicate the TDS provides a usable input for the control of household devices, as each of the subjects was able to complete all of the required tasks given to them. These results indicate the TDS can be used as an environ-mental control device when integrated with a device controller.

Figure 7. Totals of average action times for each user for X10 Commander using touch.

Figure 8. Totals of average action times for each user for IrDroid using touch.

Figure 9. Total number of errors occurring in each action of the x10 commander application.

Figure 10. Total number of errors occurring in each action of the IrDroid application for both input methods. Y axis is not labeled.


The experiment provides data on how future applications could be designed to take better take advantage of the TDS. The reported difficulty with the TDS was largely due to its implementation as a pointer device for the android. Wrong ac-tions were mostly attributed to accidental presses of another button located close the intended button. The result of this dif-ficulty is shown in the data by a large standard deviation in the actions that had the most error. The standard deviation also in-dicates the accuracy of the TDS system, especially in regards to users who are novices, is one of the limiting factors on the An-droid platform using a pointer method. Several methods could account for inaccuracy in pointer movement including a better designed UI with padded space between buttons to help prevent accidental clicks. Another method might be a gesture-based de-

sign where movements of the tongue would correspond directly to an action rather than the requirement of selecting an action button with the cursor. A gesture-based interface would reduce the need for the user to point using the TDS and might reduce errors. One indication of the performance of a gesture system was shown with the “Go Home” gesture action performed in the experiment in which subjects had no errors while also having performed closely to touch input with some cases being better. The commercial applications used were sufficiently able to provide the intended usage, as the subjects were able to com-plete all tasks. X10 proved to be a robust system with no system errors reported during the testing. Some errors occurred dur-ing the usage of the IrDroid application that were attributed to incompatibilities between the dongle control software and the

Figure 11. Questionnaire results for: (A) X10 commander intuitiveness, (B) IrDroid intuitiveness, (C) Satisfaction with the TDS per-formance, and (D) Previous experience. All figures are rated on a scale of 1 - Completely Disagree to 5 - Completely Agree.


IR protocol used by the television. Because of this error, some actions required repetition of some commands in order for the television to respond to the command. The IR dongle error was repeatedly shown in the “television on” command in which it took several repeated actions to complete the action. In addition, the television had some delay in the reception of the command and the completion of the powering on sequence, which further confused subjects resulting in multiple clicks. Multiple clicks due to no lack of knowledge of the system, however, could be mitigated in environments where the subject would have some knowledge of the actions of their television and would compen-sate for these types of delays. Results obtained about the accuracy of the TDS, and the feed-back from the subjects will be used to develop a custom applica-tion for environmental control. Further work will help refine the input to make the TDS more intuitive in the control of house-hold devices. The gesture support of the TDS will also be ana-lyzed and other gesture commands developed. This research is part of the ongoing project to develop the TDS as a more usable input method.

Acknowledgements The authors would like to thank the Georgia Institute of Technology President’s Undergraduate Research Award (PURA) program for funding the undergraduate author. Research was done in conjunction with a graduate mentor, Abner Ayala-Ace-vedo, under the supervision of Professor Maysam Ghovanloo. The TDS project was supported in part by the National Institute of Biomedical Imaging and Bioengineering grant 1RC1EB010915 and the National Science Foundation awards CBET-0828882 and IIS-0803184.

References1. Barea, R., Boquete, L., Mazo, M., & Lopez, E. (2002). Sys-tem for assisted mobility using eye movements based on elec-trooculography. IEEE Trans Neural Syst Rehabil Eng, 10(4), 209-218. doi: 10.1109/tnsre.2002.806829 2. Barreto, A. B., Scargle, S. D., & Adjouadi, M. (2000). A practical EMG-based human-computer interface for users with motor disabilities. J Rehabil Res Dev, 37(1), 53-63. 3. Cole, A. M., & Tran, B. Q. (2002). Home automation to promote independent living in elderly populations. 24th Annual Conference and the Annual Fall Meeting of the Bio-

medical Engineering Society EMBS/BMES Conference, 2002. Proceedings of the Second Joint 3, 2423 vol.2423. doi: 10.1109/IEMBS.2002.10533554. Craig, A., Tran, Y., McIsaac, P., & Boord, P. (2005). The efficacy and benefits of environmental control systems for the severely disabled. Med Sci Monit, 11(1), RA32-39. 5. Hawley, M., Enderby, P., Green, P., Cunningham, S., Brownsell, S., Carmichael, J, & Palmer, R.. (2007). A speech-controlled environmental control system for people with severe dysarthria. Medical Engineering & Physics, 29(5), 586-593. doi: http://dx.doi.org/10.1016/j.medengphy.2006.06.0096. Huo, X., & Ghovanloo, M. (2010). Evaluation of a wire-less wearable tongue-computer interface by individuals with high-level spinal cord injuries. J Neural Eng, 7(2), 26008. doi: 10.1088/1741-2560/7/2/0260087. IrDroid. Retrieved March 24, 2013, from http://www.irdroid.com8. Kim, J., Huo, X., Minocha, J., Holbrook, J., Laumann, A., & Ghovanloo, M. (2012). Evaluation of a Smartphone Plat-form as a Wireless Interface Between Tongue Drive System and Electric-Powered Wheelchairs. IEEE Transactions on Biomedi-cal Engineering, 59(6), 1796. 9. Karmali, F., Polak, M., & Kostov, A. (2000). Environ-mental control by a brain-computer interface. Engineering in Medicine and Biology Society, 2000. Proceedings of the 22nd Annual International Conference of the IEEE, 4, 2992 vol.2994. 10. Tao, C., Zhang, X., & Wang, X. (2008). Research of Envi-ronmental Control systems for Disabled people. Paper presented at the APCMBE, Beijing, China. Abstract retreived from http://link.springer.com/chapter/10.1007%2F978-3-540-79039-6_11911. X10.com. Retrieved March 24, 2013, from http://www.x10.com/homepage.htm12. X10 Commander. Retrieved March 24, 2013, from http://melloware.com/x10commander/13. Xueliang, H., Wang, J., & Ghovanloo, M. (2008). A Mag-neto-Inductive Sensor Based Wireless Tongue-Computer Inter-face. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 16(5), 504. doi: 10.1109/TNSRE.2008.200337514. Yousefi, B., Huo x., Veledar, E. & Ghovanloo, M., (2011). Quantitative and comparative assessment of learning in a tongue-operated computer input device. IEEE Transactions on Information Technology Biomed. 16(4), 633-643. doi: 10.1109/TITB.2011.2158608


ARTICLES

E f f e c t o f M a g n e t o - H y d r o d y n a m i c s a n d R a d i a t i o n Pr e s s u r e o n t h e R e i o n i z a t i o n o f D w a r f G a l a x i e sM a r t i n H a l i c e k 1, a n d J o h n Wi s e 1, P h D1 C e n t e r f o r R e l a t i v i s t i c A s t r o p h y s i c s , S c h o o l o f P h y s i c s , C o l l e g e o f S c i e n c e s , G e o r g i a I n s t i t u t e o f Te c h n o l o g y

INTRODUCTION Dwarf galaxies are born from gas clouds that coalesce in as-sociation with dark matter halos to form small galaxies that con-tain several billion stars. In comparison, the Milky Way galaxy contains approximately 300 billion stars. These galaxies are im-portant for us to study because they can elucidate the physical laws that facilitated the creation of our universe and ultimately determine the fate of our universe. Dwarf galaxies, in particu-lar, are useful tools for investigation because in depth study of a large galaxy would be too computationally expensive. Dwarf galaxies allow examination of the physical processes involved without the drawbacks of attempting to study it on the large scale. One major physical process involved in the formation of dwarf galaxies is radiation pressure from stellar radiation dur-ing the reionization phase. Radiation pressure is the pressure exerted on a particular surface by forms of electromagnetic ra-diation. Reionization refers to the stage in the creation of the universe where the predominantly neutral intergalactic medi-um is reionized by various luminous sources, including primor-dial galaxies; the period of reionization spans roughly from 100 million years to approximately 1 billion years after the Big Bang

(“Epoch of Reionization,” MIT). Without radiation pressure, a galaxy would produce a dense, metallic, stellar region due to overcooling, which is characteristic of a flaw in simulation meth-odology (Wise et al., 2012). Another important factor involved with radiation is magne-to-hydrodynamics, defined as the study of dynamic electrically conducting fluids. Moving magnetic fields throughout fluid media can create electric currents that would directly affect the amount of radiation pressure and thus the reionization. The di-rect effects of magneto-hydrodynamics on galaxy formations are still not known. Therefore, the purpose of this investigation is to create a realistic model of dwarf galaxy formations, com-paring a simulation including magneto-hydrodynamics against a control simulation involving radiation pressure. Particularly, two critical aspects of galaxy formation are magnitude of lumi-nosity and star formation rate, which will serve to determine the overall effects of magneto-hydrodynamics on galaxy formation.

METHODS The investigation was conducted entirely through compu-tational astrophysical techniques, which included running a simulation with a defined set of physical laws involved. The di-

ABSTRACTDuring the reionization period, from 100 million years to 1 billion years after the Big Bang, radiation pressure and magneto-hy-drodynamics helped shape the primordial dwarf galaxies that ionized the neutral intergalactic medium. Dwarf galaxies are impor-tant to study because they help understand the formation and behavior of early galaxies in our universe. Through computational techniques, astrophysical simulations were created to model the formation of primordial galaxies, which were investigated to produce syn-thetic observations by numerical analysis. The brightness of dwarf galaxies was calculated and compared to observed data obtained from the Hubble Space Telescope. The star formation rate was found to be regulated by magneto-hydrodynamics, in accordance with increased ion-ization fraction and clumping factor. These parameters are thought to have contributed to the enhanced luminosities of the dwarf galaxies.


rect physical laws can be modified, including accepted ones like gravity and excluding controversial or theoretical laws. Specifi-cally for this investigation, magneto-hydrodynamics was added to a simulation containing the accepted physical laws along with radiation pressure. Therefore, it can be observed how the simulation will run in different cases with different physics as-sociated. The simulations were produced by the adaptive mesh refinement code, ENZO, which is designed for advanced multi-physics dynamic astrophysics calculations and simulations. The technique used by ENZO to produce the data implements an adaptive ray-tracing scheme, shooting electromagnetic rays throughout simulated particles and gases, which comprise the created galaxy (Wise et al., 2010). A parameter file that included the necessary physics for the simulation was used to initiate the simulation. The simulations were run for a couple months each on NASA’s supercomputing cluster, Pleiades. The first simula-tion run included radiation pressure, which we will refer to as SG256-RP. The second simulation run included magneto-hydro-dynamics in addition to radiation pressure, which we will call SG256-MHD. The purpose was to investigate the effect of the

inclusion of magneto-hydrodynamics on the formation of galax-ies. Both simulations were 1 mega-parsec or 3.26 million light years co-moving in the expanding frame of the universe. SG256-RP was run for 750 million years, and SG256-MHD was run for 570 million years in the reference frame of the simulation. The length of the simulation was determined by the availability of computing time at NASA. The simulations each contained about 30 dwarf galaxies, which is approximately one thousand times smaller than the Milky Way galaxy. The simulations were analyzed using the yt-project to pro-duce synthetic visualizations of the data. Scripts were written using the yt module in Python to evaluate luminosity, star for-mation rate, and clumping factor and ionization fraction. Star formation rate was calculated by the number of stars that were created amongst all the galaxies within the respective simulation over time. Aggregate luminosity was determined by summing the brightness amongst all the galaxies within a simulation; the brightness of each galaxy was calculated within a sphere of a given radius depending on the size of the galaxy. Ionization fraction was calculated by dividing the ionized hydrogen by the

Figure 1. Histogram representing magnitude of luminosity mea-sured at 1500 Angstroms (Å) for dwarf galaxies with redshift val-ues less than nine. Hubble data obtained from telescopic obser-vations was plotted in black (Oesch et al., 2012). The galaxies were chosen at redshift less than nine because galaxies within this time region are comparable to the data obtained by the Hub-ble telescope. The data obtained from SG256-MHD are plotted in blue, and data from SG256-RP are plotted in red.

Figure 2. Plot of star formation rate plotted against time in years.


total hydrogen. Clumping factor was calculated by the average density squared divided by the square of average density. Geor-gia Institute of Technology’s computer cluster PACE was used to process the computer scripts.

RESULTS

Luminosity It was previously thought that not many dwarf galaxies formed in the development of the universe, and that they do not contribute very much to the total luminosity of the universe. However, the data suggest that there are a large number of dwarf galaxies that tend to be instrumental in the visible lumi-nosity. The data obtained from the Hubble Space Telescope is on the lower end of the intensity of luminosity, spanning from -17 to -19 (Oesch et al., 2012). For comparison, the absolute magni-tude of a Type 1A supernova is -19.3, which is approximately 5 billion times brighter than our sun (Hillebrandt et al., 2000). This is due to the fact that the Hubble Space Telescope does not have the sensitivity to detect distant dwarf galaxies. However, the simulation can model the stellar radiation from galaxies within the simulation by radiation transport. The impact of the dwarf galaxies is shown to be very dramatic on the galaxy luminosity function (Figure 1). There appears to be a maximum in the lu-minosity function around -9, after which the function seems to plateau and begin decay. It was predicted that there would be a turnover point, and the magnitude of luminosity, ϕ(M), would decay according to a power law,

ϕ(M) dM = (2/5) ϕ* ln(10)[10(2/5) (M* - M)]α+1 exp[-10(2/5)(M* - M)] dM (1)

where M is magnitude and ϕ*, α, and M* are determined em-pirically (Simard, 1996). The luminosity data from SG256-MHD fit the prediction of the turnover point described by equation (1), and henceforth indicate that inclusion of magneto-hydrody-namics may be a better representation of the physics involved in galaxy formation compared to the simulation with all factors except magneto-hydrodynamics, SG256-RP.

Star Formation Rate The star formation rate dictates the number of stars that are being formed per unit time. The simulation run with magneto-hydrodynamics (SG256-MHD) had a greater star formation rate at an earlier time compared to SG256-RP. Therefore, the data suggest that magneto-hydrodynamics stimulate star formation rate, because the largest star-forming time region occurs at 425 million years after creation in SG256-MHD compared to 525 mil-lion years in SG256-RP. It is known that in a galaxy below 40 K most energy is in the form of magnetic fields. These are the conditions at which these stars are forming, so it is reasonable that magnetic fields can differently affect star formation rates by increasing the amount of energy contained (Price et al., 2010).

Figure 3. Ionization fraction (3A) and clumping factor (3B) are plotted against time with the same x-axis ranging from 200 mil-lion years to 800 million years.


Clumping Factor / Ionization Fraction The ionization fraction, defined as the ratio of ionized hy-drogen to total hydrogen, greatly increases in SG256-MHD compared to SG256-RP (Figure 3A). The ionization fraction increases exponentially and maximizes in SG256-MHD at 550 million years, compared to SG256-RP, which does not reach this level until it gradually increases at 750 million years. The clump-ing factor, defined as the average of density squared over the square of average density, in SG256-MHD is similar to SG256-RP (Figure 3B). A higher resolution is necessary to clarify the impli-cations of the lack of differentiation between clumping factors rates between SG256-RP and MHD. The recombination rate of hydrogen is directly proportional to the clumping factor. The in-creased ionization fraction is a direct consequence of the higher star formation rates seen in SG256-MHD.

DISCUSSION The purpose of this investigation was to create a realistic model of dwarf galaxy formations, comparing a simulation in-cluding magneto-hydrodynamics against a control simulation involving radiation pressure. Using the magnitude of luminos-ity and star formation rate to determine the overall effects of magneto-hydrodynamics on galaxy formation, it was hypoth-esized that inclusion of magneto-hydrodynamics in simulation would produce more realistic results, fitting the data obtained from the Hubble Space Telescope. Recent literature suggests that radiation pressure must be included in astrophysical simu-lations to avoid creating a super-dense, highly metallic region, which is erroneous. These investigations reached this conclu-sion by constructing two simulations using radiation pressure as a dependent variable (Wise et. al, 2012). Using a comparable technique in this investigation, magneto-hydrodynamics was used a variable compared to a control to establish its presence in astrophysical phenomena. The stronghold of the data, the fit of the luminosity function, indicates that the simulations run with the inclusion of magneto-hydrodynamics correspond well with the intensity of brightness observed in our own night sky when compared to observed galaxies obtained by the Hubble Space Telescope (Oesch et al., 2012). The results show that magneto-hydrodynamics increases the rate of star formation. The greatest star-forming time region oc-curs at 425 million in SG256-MHD compared to 525 million years in SG256-RP. There is a greater ionization of hydrogen due to the inclusion of magneto-hydrodynamics. The ionization fraction increases exponentially and maximizes in SG256-MHD at 550 million years, compared to SG256-RP, which maximizes at 750 million years. Ionization seems to increase star formation, which would explain the brighter luminosity observed that diminishes

after the maximum. However, the results of the clumping factor appear to disagree with these findings. There is not a notice-able increase in clumping factor with the addition of magneto-hydrodynamics into the simulation. This seems to conflict with the increases in ionization fraction and star formation rate, but it may be possible that ionization fraction increases the star forma-tion rate enough that an increase in clumping factor is not neces-sary for the appropriate luminosity function to be produced. All things considered, the inclusion of magneto-hydrodynamics ap-pears to be consistent with the physical observations of known dwarf galaxies. As more computational research is conducted, a better picture of our universe is constructed, allowing us to further understand our cosmological origins and fate. Subsequent research projects could be conducted by adding more resolution to the simulations, along with expanding the simulation timeframe, which would produce more thorough re-sults. The time that the simulations are run depends on resourc-es available. In Figure 1, it is possible that for the more luminous dwarf galaxies, one star or a small collection of stars could have contained the majority of the flux observed. Examination of the data obtained in the simulation indicates the presence of over-dense highly luminous areas, which suggest that this might have occurred. Increasing resolution, which is similar to increas-ing resolution on a television, combats this because it allows for greater differentiation and definition. The predicted turnover point of the graph for the luminosity magnitude at 1500 Å could not be seen in the research conducted, but if the resolution were increased the turnover point could be accurately determined and this would eliminate the likelihood of a small group of stars over-representing the luminosity of a dwarf galaxy. The resolu-tion available also depends on resources, as increasing will ex-tend the time that the simulations are run by a couple months. Further experimentation into this subject is necessary to accu-rately discern the picture of our expanding universe. Using tools such as luminosity functions and star formation rates, we can determine that we have created accurate simulations of galaxies. This is essential to help us unravel the fate of the universe we live in and how galaxies behave over time.

Acknowledgements I would like to thank my research advisor, Dr. John Wise, for his assistance in running the simulations and his aid in writing the code to analyze the data obtained. This project would not be possible without his guidance, and it is very much appreciated.


References1. Hillebrandt, W. and J.C. Niemeyer. “Type IA Superno-va explosion models.” Annual Review of Astronomy and Astro-physics 38 (2000): 191-230.2. Oesch, P A et al. “The Bright End of the UV Luminosity Function at z ∼ 8: New Constraints from Candels Data in Goods-south 1.” Astrophys J, 759 (2012): 135–48. Print.3. Price, Daniel J, and Matthew R Bate. “Magnetic Fields and Radiative Feedback in the Star Formation Process.” AIP Conf. Proc. 1242 (2010): 205–18. Print.4. Simard, Luc. “The Local Galaxy Luminosity Function.” U of Cal Observatories. 2 Sep. 1996. Web. 27 Feb. 2013. http://www.ucolick.org/~simard/phd/root/node7.html .5. Turk, M. J. et al. “yt: A Multi-code Analysis Toolkit for

Astrophysical Simulation Data.” Astrophys J Sup 192 (2011): 9. Print.6. Wise, John H, and Tom Abel. “Enzo + Moray : Radiation Hydrodynamics Adaptive Mesh Refinement Simulations with Adaptive Ray Tracing.” Mon. Not. Roy. Astron. Soc 414 (2011): 3458–91. Print.7. Wise, John H et al. “The Birth of a Galaxy . II . The Role of Radiation Pressure.” 8. Mon. Not. Roy. Astron. Soc 427 (2012): 311–28. Print. 9. “Epoch of Reionization.” MIT Haystack Observatories. Web. 27 Feb. 2013. http://www.haystack.mit.edu/ast/science/ep-och/index.html.


ARTICLES

ABSTRACTThe U Visa is an immigration visa designed by U.S. Citizenship and Immigration Services (USCIS) and the Department of Homeland Security (DHS) to encourage undocumented immigrant victims of criminal acts occurring within the United States to report to the police, to participate in criminal investigations as witnesses, and to obtain a legal immigration status through a work authorization card and pathway to permanent residency due to their cooperation with law enforcement. The public purpose of the U Visa policy is to create and enhance the trust between law enforcement officers and the undocumented immigrant community. The current reality is that there is a severe lack of reporting of violent crimes when undocu-mented immigrants are victimized. One of the critical steps to preparing a U Visa application to submit for USCIS consideration is a document called the certification form, which is filled out by a local law enforcement agency that has information regarding the helpfulness of the immigrant victim with regards to the investigation of the criminal act or case. This review explores some of the barriers that law enforcement agencies have put into place, whether it is through systemic policies that limit the number of cases that they actually deem eligible for certification, denials based on unjus-tified grounds built from agency level discretion, or an outright refusal to develop the capacity of their agency to handle such certification requests.

M i c h a e l K . C r i n o 1 I I I a n d Va l e r i e A n n We r n e r 2, P h D, L M F T1 D e p a r t m e n t o f U r b a n P l a n n i n g a n d Po l i c y, C o l l e g e o f U r b a n P l a n n i n g a n d Pu b l i c A f f a i r s , U n i v e r s i t y o f I l l i n o i s a t C h i c a g o 2 D e p a r t m e n t o f Le a d e r s h i p a n d Pu b l i c S e r v i c e , A d l e r S c h o o l o f Pr o f e s s i o n a l Ps y c h o l o g y, U n i v e r s i t y o f I l l i n i o s a t C h i c a g o

A r t i c l e E xc h a n g ew i t h t h e U n i v e r s i t y o f I l l i n o i s a t C h i c a g o

U n i t e d S t a t e s C i t i z e n s h i p a n d I m m i g r a t i o n S e r v i c e s U V i s a C e r t i f i c a t i o n D e n i a l b y L a w E n f o r c e m e n t : A n O b s t a c l e f o r I m m i g r a n t V i c t i m s o f C r i m e


INTRODUCTION The U Visa is an immigration visa designed by U.S. Citizen-ship and Immigration Services (USCIS) and the Department of Homeland Security (DHS) to encourage undocumented immi-grant victims of criminal acts occurring within the United States to report to the police, to participate in criminal investigations as witnesses, and to obtain a legal immigration status through a work authorization card and pathway to permanent residen-cy due to their cooperation with law enforcement. Created by the Victims of Trafficking and Violence Protection Act of 2000, the U-visa was designed to provide legal immigration status to noncitizens who are assisting or are willing to assist authorities investigating crimes. There are four basic categories that USCIS established to determine whether or not an undocumented im-migrant victim may be eligible for the U Visa; (1) has been a victim of one of the violent crimes as specified by USCIS; (2) has suffered substantial physical or mental abuse from the qualify-ing criminal activity; (3) has information regarding the criminal activity; (4) has assisted government officials in the investigation or prosecution of such criminal activity; and (5) the criminal ac-tivity violated US law or occurred in the United States (includ-ing Indian country and military installations) or the territories and possession of the United States (Kinoshita, 2009). The public purpose of the U Visa policy, as specified by US-CIS, is to create and enhance the trust between law enforcement officers and the undocumented immigrant community. The cur-rent reality is that there is a severe lack of reporting of violent crimes when undocumented immigrants are victimized. The lack of the reporting of violent crime is due to a fear of identifi-cation by the U.S. government or local law enforcement, which leads many immigrants to fear deportation. Because of the per-petration of domestic violence and sexual assault within immi-grant communities, an overwhelming majority of U Visa appli-cants are women and children. This is due to the perpetration of domestic violence and sexual assault within immigrant commu-nities (Rana, 2012). The U Visa program serves as a public tool to embolden women victims, and gives them the confidence to not fear legal consequences if they report and prosecute an intimate partner (Public Law 106–386, 106th Congress, 2000). The U Visa serves the humanitarian policies of the United States by guaranteeing the legal safety of individuals who were victimized by crime and assisted in the investigation of criminal acts. The U Visa, by granting legal status through the U nonim-migrant status work authorization card, helps seriously harmed victims recover and rebuild their lives. In addition, by empow-ering undocumented victims to report criminal acts, it removes the impunity from the law that perpetrators have traditionally experienced in immigrant communities and overall making

communities more permeable to law enforcement, and increases public safety (U.S. Citizenship and Immigration Services, 2010). It encourages the cooperation of a particular victim in giving helpful testimony to a criminal act by giving the access to a le-gal immigration status. In sum, the U Visa removes many of the barriers that undocumented victims experience, such as fear of their immigration status, and insecurity about complying with law enforcement (Ibid, 2010). One of the critical steps to preparing a U Visa application to submit for USCIS consideration is a document called the cer-tification form, which is filled out by a local law enforcement agency that has information regarding the helpfulness of the im-migrant victim with regards to the investigation of the criminal act or case. In many cases, it is the local police department who acts as the certifying agency; or if criminal charges were pressed in the county court, then it would be the prosecutor or judge of the criminal court who could testify to the victim’s helpfulness (Department of Homeland Security, 2007). This project explores some of the barriers that law enforcement agencies put into place, whether it is through systemic policies that limit the num-ber of cases that are deemed eligible for certification, denials based on unjustified grounds built from agency level discretion, or an outright refusal to develop the capacity of their agency to handle such certification requests.

Domestic Violence and other Crimes in the Latino Immigrant Community Domestic violence is a universal problem affecting commu-nities world-wide and forms part of societal institutions in com-munities where men have the dominant role in the society, in the family, and in romantic relationships. The Latino immigrant community within the United States is an ethnic mix of cultures and different Spanish linguistic characteristics from countries such as Mexico, Guatemala, El Salvador, Honduras, Panama, Costa Rica, and Caribbean countries, such as the Dominican Republic, Puerto Rico, and Cuba. Although there are many re-gional and linguistic differences, there is a uniting cultural char-acteristic of machismo among many men who constitute part of the lower working class (Walton-Moss, 2002). Domestic violence is more prevalent in some immigrant communities and populations than in others due to its legal and social acceptance within the local context in addition to how the roles of men and women are defined by the culture of origin. Within the Latino immigrant community, domestic violence in many cases is institutionalized and tolerated by society, as it is passed down through the dominant male patriarchal head of the house. In addition, a male dominated Catholic religious sys-tem in which divorce is highly discouraged and the woman is


taught to “accept” the suffering just as part of her experience as being a wife, mother, and woman makes the position of a fe-male victim difficult to change (Dobash, 1978). It is important to recognize, that in a historical context, the woman is the “appro-priate victim”; as opposed to being the “inappropriate” victim. Thus, “wife beating”, or abuse of the woman spouse, has been traditionally accepted in many communities without access to higher education and with few female-oriented economic op-portunities. In a male dominated society, intimate partner abuse is purposeful for the maintenance, enforcement, and exercise of power, control and authority as embraced by the traditional roles of men in undereducated, under-resourced communities (Ibid. 1978).

Certification Form Support by Law Enforcement In order for the U Visa application to have merit for USCIS consideration, it must have the accompanying form I-918B, which is a form that a law enforcement official signs verify-ing the actual “helpfulness” of the victim to the investigation or prosecution of the criminal act. Since USCIS issued the of-ficial guidelines for the U Visa policy in 2007, there have been a substantial number of law enforcement agencies who received U Visa certification trainings, and were required to review the updated policies through official USCIS communications. Ac-cording to USCIS, the U Visa policy could be a tool used to pro-mote public safety in immigrant communities and to encourage victims of crime to step forward and identify the perpetrators and help with the investigation, prosecution, and deportation of violent offenders. Urban areas with high immigrant popu-lations have been receptive to the U Visa policy after being al-lowed sufficient time to review, understand, and train person-nel in regards to the U Visa certification process which involves a review of the helpfulness of victims on a case-by-case basis. Since 2009, over 25,000 U Visa cases (not counting eligible fam-ily members) have been approved by USCIS and work autho-rization cards were granted to the victims and qualifying fam-ily members (U.S. Citizenship and Immigration Services, 2011). However, despite the high numbers of cases submitted and the U Visa granted to recipients, there have been incidents of reluc-tance to sign the certification form, whether at a local level with a municipal police agency, or county level with the state’s attor-ney office (Wallace, 2012). There is noted non-compliance with the spirit of the U Visa policy as defined by USCIS as being a tool to which law enforcement in efforts to deter crime. It allows law enforcement to better access the immigrant community in regards to investigating crime and also that victims who have cooperated and proved helpfulness are able to receive an immi-gration relief on humanitarian grounds.

Misconception of the U Visa at the Local Level In the case of the New York Police Department, 125 U Visa certifications have been signed, which constitutes 56% of the to-tal certification requests received by the department. In contrast, the Oakland Police Department in California has signed 1,201 certification forms since 2009, which is 95% of the certification requests received (Glorioso and Burke, 2012). It is important to note that, due to agency level discretion and overall perceived importance of the U Visa policy in regards to concerns for public safety in immigrant communities, many counties fully embrace the policy, some embrace it yet do not prioritize it, and some counties in Illinois, Virginia, Connecticut, and Maryland have established systemic policies that severely limit the nature of the U Visa (Haufreck, 2012). In the case of Illinois, the city of Chicago in Cook County has been receptive to the U Visa policy and has granted a total of 384 certifications for the U Visa since 2009, with five being granted in 2009 and over 155 in the year 2012 (Wallace, 2012). This embrac-ing of the U Visa policy, as mentioned by advocates and the evi-dence of the sharp increase in the number of certification forms granted to victims may be due to the significant population of undocumented immigrants and high incidence of inter-partner violence in the concentrated immigrant communities. In the case of Dupage County, which lies in the suburban areas of the met-ropolitan region of Chicago and Northern Illinois, obtaining cer-tification documents for eligible victims has proven to be chal-lenging as there may exist a misunderstanding of the purpose of the U Visa. In addition, the recent evolving ethnic dynamics of the suburban communities has perhaps added some misconcep-tions of the Latino immigrant community. The States Attorney’s Office in Dupage County related that they have “the freedom to decide to approve the certifications or deny them” (Kahler, 2012), but do not provide real USCIS supported reason for the denial except for capacity purposes or “agency discretion” for the certification denial.

Identified Reasons for Certification Denial These denial policies that are formed by law enforcement agencies are legal in nature, due to the discretion that USCIS has given the certification agency(s) to establish for itself admin-istrative roles and an autonomous review process. However, many law enforcement agencies or prosecutors’ offices choose to enact administrative policies that do not serve the public pur-poses of the U Visa policy in granting the U Visa as a justice measure due to the victimization of the immigrant (Abrams, 2010). In addition, many of the denial policies utilized by law enforcement agencies clearly work against and do not follow the clear language of the USCIS guidelines. Whether or not these


controversial policies are enacted due to an anti-immigrant so-cial sentiment or not, the reasons identified for certification de-nial include the following: a certification statute of limitations; absence of court appearances; failure to testify in court; lack of an offender conviction; victim still lives with the abuser; abuser was not deported by ICE; certification only granted for victims in current investigations; absence of criminal charges in police report follow-up; lack of substantial emotional or physical harm; lack of organizational resources; criminal activity of the victim themselves; and a decision to not deal with the U Visa policy for political reasons. It is important to note that the certification agency’s decision to review and provide a signature regarding victim helpfulness is “entirely discretionary… the agency is un-der no legal obligation to complete a Form I-918, Supplement B, for any particular alien” (Jenson, 2009). However, without the certification form, the immigrant victim is ineligible to apply for the U nonimmigrant status through USCIS. Due to the current administrative freedoms granted to local law enforcement by USCIS, the U petition has been underuti-lized and will continue to be underutilized without an effective certification process. Due to the underutilization of the U Visa, law enforcement is not able to enhance its permeability to immi-grant communities through the investigation of crime, and the policy is not able to assist victims who have cooperated with law enforcement. Without the certification document USCIS is unable to appropriately evaluate the helpfulness of the victim and the validity of the crime, and thus it positions law enforce-ment agencies as the “gate-keeper” to U Visa relief for victims (Abrams, 2010). Due to the increasing role of law enforcement engaging in strong immigration policies with programs such as Safe Communities, in which Immigration and Customs Enforce-ment has streamlined case processing with local law enforce-ment, agencies are many times placed under pressure to assist undocumented victims of crime while still enforcing tough im-migration policies. Since the certification step to the U Visa ap-plication falls on law enforcement to decide whether or not they have the organizational will and capacity to work with a vic-tim of crime and verify the helpfulness of the immigrant, many times the practice of the policy is assigned a low priority due to the time-cost of the review process.

CONCLUSIONS The U Visa policy serves the public interests in of the United States by increasing public safety, permeability of immigrant communities to law enforcement in undocumented immigrant communities, and the reporting and prosecution of domestic violence and sexual assault affecting immigrant women. The hesitancy law enforcement agencies exhibit to perform the certi-

fication component of the application does not serve the public interests of the United States. This is because it discourages the use of the U Visa as a tool to combat crime and serve the human-itarian purposes of USCIS. Ultimately, the policy adoption and implementation at all levels - federal, state, and local - is about the justice measure granted to the victim due to the helpfulness demonstrated in the prosecution of the crime. There is an in-crease in public awareness and public safety as abusers realize that their potential immigrant victims have been empowered by civil rights to not tolerate manipulation, control, and abuse. It is a human right to be able to report crime, to cooperate with law enforcement in the investigation of the crime, to aid in the prosecution of the criminal act, and to receive an immigration adjustment for purposes of legal protection and humanitarian interests, which is a public priority of the United States govern-ment.

References1. Abrams, J. (2010). The Dual Purposes of the U Visa Thwarted in a Legislative Duel. Saint Louis University Pub-lic Law Review Vol. XXIX: 373. Saint Louis University School of Law. Retrieved on December 23rd, 2012. Available at: http://www.slu.edu/Documents/law/PLR/Archives/PLR29-2_Abrams_Article.pdf2. Department of Homeland Security. (2007). New Clas-sifications for Victims of Criminal Activity; Eligibility for “U” Nonimmigrant Status. [CIS No. 2170–05; DHS Docket No. US-CIS– 2006–0069] RIN 1615–AA67. Federal Register/ Volume 72, No. 179 / Monday, September 17, 2007 / Rules and regulations. Pages 52014-53042. Retrieved on December 23rd, 2012. Avail-able at: http://apps.americanbar.org/domviol/tip/trainings/Im-migration%20Remedies%20for%20Trafficking%20Victims%20Workshop/U%20Visa%20Regs%20-%20Fed.%20Register%209.17.2007.pdf3. Department of Homeland Security. U.S. Citizenship and Immigration Services.United States of America. (2011) Form I-914 – Application for T Nonimmigrant Status and Form I-918 – Petition for U Nonimmigrant Status, Visa U. Service-wide Re-ceipts, approvals, and denials. November 3, 2011. Performance Analysis System (PAS), Office of Performance and Quality (OPQ), Data Analysis and Reporting Division. Fiscal Years 2002-2011. 4. Dobash, R.E. Dobash, R. (1978). P. 12: Wives: The “Ap-propriate” Victims of Marital Violence (1978). Reflection. PART TWO / Physical Violence against Women. 184-186. Violence Against Women. Classic Papers edited by Raquel Kennedy Bergen, Jeffrey L. Edleson and Claire M. Renzetti. Retrieved


on December 23rd, 2012. Available at: http://www.mincava.umn.edu/classics/chapters/Chapter%2012%20Dobash%20&%20Dobash%20reflection%20by%20Dobash.pdf5. Glorioso, C. Burke, T (2012). NYPD Reluctant to Sign U-Visa Letters. Critics say officials are ignoring an important crime-fighting tool. Jun 15, 2012. I –Team, Copyright 2012. Re-trieved on December 18th, 2012. Available at: http://www.nbc-newyork.com/news/local/I-Team-U-Visa-Illegal-Immigrants-NYPD-Crime-Fighting-157547355.html6. Haufreck, Mark. (2012). Telephone Interview. Adminis-trative Director for Mil Mujeres, Washington, D.C. Conducted on December 18th, 6:40 pm. 7. Jensen, T. (2009). U Visa “Certification”: Overcoming the local hurdle in response to a federal statute. Idaho Law Re-view, 2009 45 Idaho L. Rev. 691. Available through Lexis Nexis, through student affiliation through the University of Illinois at Chicago Richard M. Daley Library, UIC. 8. Kahler, K. (2012). A birthday blessing: An abused wom-an’s journey to receive her U Visa. 2010, Immigrant Connect. Retrieved on December 18th, 2012. Available at: http://www.im-migrantconnect.org/2012/12/09/a-birthday-blessing-one-wom-ans-journey-to-receive-her-u-visa/9. Kinoshita, S. (2009). The U Visa. Obtaining Status for Immigrant Victims of Crime. 1st edition. The Immigrant Legal Resource Center. Copyright 2009. Immigrant Legal Resource Center. 1663 Mission Street, Suite 602, San Francisco, CA 94103. 10. Public Law 106–386. 106th Congress Oct. 28, 2000 [H.R. 3244]. Victims of Trafficking and Violence Protection Act of 2000. Title V – Battered Immigrant Women- Sec. 1513. Protec-tion for certain crime victims including victims of crimes against women- Public Law 106–386, 114 Stat. 1464—Oct. 28, 2000 114 Stat. 1464- 1548. Retrieved on January 2nd, 2013. Available at: http://www.state.gov/documents/organization/10492.pdf

11. Rana, Sheetal. (2012) Addressing Domestic Violence in Immigrant Communities: Critical Issues for Culturally Com-petent Services. (2012, February) Harrisburg, PA: VAWnet, a project of the National Resource Center on Domestic Violence. National Resource Center on Domestic Violence. All rights re-served. 3605 Vartan Way, Suite 101, Harrisburg, PA 17110. Re-trieved on December 18th, 2012. Available at: http://vawnet.org/print-document.php?doc_id=3157&find_type=web_desc_AR12. U.S. Citizenship and Immigration Services. (2010). Questions and Answers about U Nonimmigrant Status (aka the “U Visa”). Last updated: 11/22/2010. Retrieved on December 21st, 2012. Available at: http://www.uscis.gov/portal/site/uscis/menuitem.5af9bb95919f35e66f614176543f6d1a/?vgnextoid=1b15306f31534210VgnVCM100000082ca60aRCRD&vgnextchannel=ee1e3e4d77d73210VgnVCM100000082ca60aRCRD13. Wallace, A. (2012). As demand for U Visa grows, visa’s future for abused women uncertain. December, 2012. In Fear-ing the Law, Problems with Papers, What’s New. Immigrant Connect Copyright © 2010. All rights reserved. Retrieved on December 18th, 2012. Available at: http://www.immigrantcon-nect.org/2012/12/09/as-demand-for-u-visa-grows-visas-future-uncertain/14. Walton-Moss, B., Campbell, J. (2002). Intimate Partner Violence: Implications for Nursing. ANA Periodicals OJIN Ta-ble of Contents Volume 7 - 2002. No 1: Jan 02 Intimate Partner Violence. © 2002 Online Journal of Issues in Nursing Article published January 31, 2002. Retrieved on December 28th, 2013. Available at: http://www.nursingworld.org/MainMenuCatego-ries/ANAMarketplace/ANAPeriodicals/OJIN/TableofContents/Volume72002/No1Jan2002/IntimatePartnerViolence.html


We are dedicated to enhancing the health and well-being of animals.

Merial, a world leading animal health company, is a proud contributor to Georgia’s growth in the biotech sector.

GeorGeorGeorGeorGeorgia Biomagia Biomagia Biomagia Biomagia Biomass rss rss rss rss recececececognizognizognizognizognizeeeees and prs and prs and prs and prs and proudoudoudoudoudly supportly supportly supportly supportly supports the acs the acs the acs the acs the acaaaaademic edemic edemic edemic edemic exxxxxcellence and innocellence and innocellence and innocellence and innocellence and innovvvvvaaaaativtivtivtivtive se se se se stttttandarandarandarandarandards of theds of theds of theds of theds of theGeorGeorGeorGeorGeorgia Tgia Tgia Tgia Tgia Tech eech eech eech eech educducducducducaaaaational stional stional stional stional syyyyyssssstttttem.em.em.em.em.

As the larAs the larAs the larAs the larAs the largggggeeeeessssst prt prt prt prt production foduction foduction foduction foduction faciaciaciaciacility of induslity of induslity of induslity of induslity of industrial fuel grtrial fuel grtrial fuel grtrial fuel grtrial fuel graaaaade wde wde wde wde wood pelleood pelleood pelleood pelleood pellettttts, Geors, Geors, Geors, Geors, Georgia Biomagia Biomagia Biomagia Biomagia Biomass usess usess usess usess uses suss suss suss suss sustttttainable,ainable,ainable,ainable,ainable,rrrrreneeneeneeneenewwwwwable naable naable naable naable naturturturturtural ral ral ral ral reeeeesoursoursoursoursourcececececes and applies and applies and applies and applies and applies as as as as advdvdvdvdvanceanceanceanceanced td td td td technology technology technology technology technology to help meeo help meeo help meeo help meeo help meet the et the et the et the et the evvvvvererererery gry gry gry gry grooooowing enerwing enerwing enerwing enerwing energy neegy neegy neegy neegy needs ofds ofds ofds ofds ofthe wthe wthe wthe wthe world. World. World. World. World. We are are are are are fe fe fe fe focuseocuseocuseocuseocused on the futurd on the futurd on the futurd on the futurd on the future. Lee. Lee. Lee. Lee. Learn morarn morarn morarn morarn more about us and re about us and re about us and re about us and re about us and reneeneeneeneenewwwwwable enerable enerable enerable enerable energy agy agy agy agy at t t t t wwwwwwwwwwwwwww.....gggggabiomaabiomaabiomaabiomaabiomass.css.css.css.css.comomomomom.....

For more information, please call 770.663.0663

Johnson Controls understands buildings better than anyone else. Our experience spans more than 120 years and tens of thousands of buildings. The owners and managers of commercial, institutional and government buildings worldwide turn to us to maximize comfort, productivity, safety and energy efficiency.

Increase net operating income by implementing whole building retrofits utilizing lifecycle energy and operational cost reductions, enabled by attractive financing options.

Making buildings more energy efficient, comfortable, safe and productive

Robert W. Allison 1938

H. Frank Smith 1950

Lanny S. Thomas 1974

Todd Grossweiler 1999

Christopher C. Reichart 2002

Mark R. Gallacher 2005

R. Brian Russ 2006

Andrew B. Norman 2006

2284 Marietta Boulevard, N.W. Atlanta, Georgia 30318 (404) 351-6430 fax (404) 350-1065

Allison-Smith Company LLC Electrical Contractors & Engineers

BBBeeesssttt

wwwiiissshhheeesss tttooo ttthhheee ssstttuuudddeeennntttsss ooofff GGGeeeooorrrgggiiiaaa TTTeeeccchhh!!!!!!

Georgia Tech graduates

of Allison-Smith Co.

BUILDING FOR GENERATIONS

www.jmwilkerson.com

1734 Sands Place

Marietta, GA 30067

Office: 770.953.2659

Fax: 770.933.9665

Oglethorpe PowerCorporationProviding Georgia’s electric membership corporations with reliable electricity for more than 35 years

Oglethorpe Power Corporation is the nation’s largest power-supply cooperative with more than $8 billion in assets, serving 39 electric membership corporations, which collectively provide electricity to 4.1 million Georgia citizens. A proponent of conscientious energy development and use, we balance reliable and affordable energy with environmental responsibility and have an outstanding record of regulatory compliance. Our diverse energy portfolio includes natural gas, hydroelectric, coal and nuclear generating plants, as well as purchased power. Headquartered in Tucker, Ga., Oglethorpe Power was established in 1974 and is owned by our 39 member systems.

www.opc.com

5036 B.U. Bowman DriveBuford, GA 30518

Phone: 770-904-4444Fax: 770-904-0888

For more information, please call 770.663.0663

Johnson Controls understands buildings better than anyone else. Our experience spans more than 120 years and tens of thousands of buildings. The owners and managers of commercial, institutional and government buildings worldwide turn to us to maximize comfort, productivity, safety and energy efficiency.

Increase net operating income by implementing whole building retrofits utilizing lifecycle energy and operational cost reductions, enabled by attractive financing options.

Making buildings more energy efficient, comfortable, safe and productive

10% OFF 10% OFF 10% OFF 10% OFF 10% OFF for studentsand faculty with ID*

*Tires, battries, and oil changesexcluded. Cannot be combined

with any other offers.