Using Remote Sensing to Measure Bridge Condition Using Remote Sensing to Measure Bridge Condition –– An Example of Collecting, Analyzing, & Visualizing An Example of Collecting, Analyzing, & Visualizing Thermal, Radar, 3Thermal, Radar, 3‐‐D Optical, & D Optical, & LiDARLiDAR Data for the Data for the 

Same Michigan BridgesSame Michigan BridgesSame Michigan BridgesSame Michigan Bridges

ASPRS Joint WGLR‐EGLR meeting 6/22/12Briefing from Project Wrap‐up Meeting

Tuesday June 19 2012Tuesday June 19, 2012Houghton, MI

1

Project TeamMichigan Technological UniversityMichigan Technological Universityg g yg g y

Tess Ahlborn, Ph.D., P.E., FPCIColin Brooks

Devin Harris, Ph.D.Robert Shuchman, Ph.D.

Larry Sutter, Ph.D.Joe Burns, Ph.D.Arthur EndsleyRick Dobson

Kiko de Melo e SilvaKhatereh Vaghefi

Renee OatsRyan HoensheidNathan Miller

Center for Automotive ResearchCenter for Automotive ResearchRichard WallaceQiang Hong

www.mtti.mtu.edu/bridgecondition/www.mtti.mtu.edu/bridgecondition/Qiang HongMike Forester

2

Outline• Motivation and Background• Technology Selection• Technology Selection• Technology Overview and Performance• Decision Support System• Economic Evaluation• Review of Goals and Performance

3

Bridge Condition in the U.S. > $150B to repair today; safety; fewer personnel, budget challenges

Deteriorated Concrete Deck

Deteriorated Bearing

Deteriorated Concrete Element

4

Deteriorated Concrete ElementSettlement Corrosion and Section Loss

Technical ApproachGoal:1. Assess the potential for commercial remote sensors . Assess the potential for commercial remote sensors

to be used to assess bridge condition and performance.

2. Deploy commercial sensors on in‐service bridges to assess condition.

3. Develop decision support system (DSS) for integrating remote sensing into current bridge 

iassessment practices.

5

Project ConceptRemote Sensing – for bridge engineers: enhanced bridge inspection at highway 

speed without traffic disruption (e.g. collecting information at a distance)

DSS

Period 0(Baseline)

Bridge Management System DataStructural Health Monitoring ModelMaintenance RecordsMeteorological Data

Transportation officials utilize dynamicid l h i l

rderR

erks rrkjkRj

ˆ22

4,

DSS: algorithms & interface

DSS

ME

Period 1

Bridge Health

 Indicators

Bridge Health Signature to evaluate changing condition

Periodic assessments enhanced with remote sensing as trouble spots are identified

Decision Support System

Trouble Spot 2Trouble Spot 1

BRIDGE MANAGEMENTTEAMBRIDGE

TIM

MANAGEMENT TEAM

On Site andIn Situ Sensors

BRIDGEData Collection

Relay

Period X(Current)Bridge Health Signature

Damage Location

6

y

Intended Impacts

• Enhance current bridge inspection practices using Commercial Remotepractices using Commercial Remote Sensing and Spatial Information

• Provide method for expedited discovery of bridge defectsg

• Evaluate cost effectiveness of remote sensing for bridge inspectionssensing for bridge inspections

7

T h l O i dT h l O i dTechnology Overview and Technology Overview and PerformancePerformance

8

Top Priorities / ChallengesTechnology Selection: commercially available technologies to enhance current inspection processes, including safety, 

Location Applicable TechnologiesS f 3 D O ti St t Vi St l Ph t h

while minimizing traffic disruption.

Surface 3‐D Optics, Street View Style Photography,  LiDAR, GigaPan

Subsurface Infrared Thermography (Thermal IR), Ultra WideSubsurface Infrared Thermography (Thermal IR), Ultra Wide Band Imaging Radar System (UWBIRS) / GPR

Global Metrics Digital Image Correlation LiDAR InterferometricGlobal Metrics Digital Image Correlation, LiDAR, Interferometric Synthetic Aperture Radar (InSAR), satellite imagery

9

Initial Sensor Evaluation and Selection

10

3D Optical Bridge Evaluation System (3DOBS) –close‐range photogrammetryg p g y

Deployment on Willow Road over US-23 during August 2011 3DOBS‐derived elevation data can detect

Visual of percent spalled area for the Willow (6.99% spalled) Road bridge using 3DOBS data as the input and ArcGIS as

g gfield demonstrations

3DOBS derived elevation data can detect spalls at various minimum sizes 

11

3DOBS data as the input and ArcGIS as the analysis software

Additional value from 3DOBS bridge deck DEM: IRI (International Roughness Index)• Extracting maximum value from high‐resfrom high res bridge DEM

• Used 3 DOBS DEM extractedDEM, extracted out points along wheel tracks, processed dataprocessed data using ProVAL

• Obtained very high res IRI data;high‐res IRI data; roughness data can help inform the condition of

12

the condition of the bridge deck

Additional value: assessing crack density• 3DOBS creates a very high 

resolution image of the bridge surface – can we use that to assess crack density?

• Investigated to demonstrate obtaining additional value out gof already collected data sets

• Characterize the cracks –– Transverse vs. longitudinalTransverse vs. longitudinal 

vs. map (pattern) cracking– Width ‐ <1/16”, 1/16‐1/8”, 

1/8‐3/16”, > 3/16”

Original crack image Image after running though MATLAB

algorithm; refining & tif i– Density ‐ Cumulative lineal 

length per 100 ft or other distance

& quantifying output.

13

BridgeViewer Remote Camera System (BVRCS) – low‐cost system

Deployed on Freer Rd to capture a bridge photo p g pinventory location of the digital photographs being

displayed in Google Earth; each box contains a hyperlink to a full-resolution view of the photo

14

yp ptaken at that location

GigaPan – Mannsiding Rd example

Profile view of Mannsiding Rd from a GigaPan image. The full resolution version of this photo captures the entire

Mannsiding Rd Bhttp://gigapan.org/gigapans/84462/Freer Rd Bridge: http://gigapan.org/gigapans/91311/

version of this photo captures the entire side of the bridge at very high resolution.

GigaPan system collecting high-resolution bridge

15

g ginventory photos

Surface: 3DOBS, BVRCS, GigaPan• Technology Features

– Low cost components, rapid deployment, limited time to collect data

– Useful metrics: % area and volume & location of spalls, International Roughness Index (IRI), geo‐tagged and very g ( ), g gg yhigh resolution photo inventory

• Limitations5mm resol able feat res ( ith c rrent set p) a tomation– 5mm resolvable features (with current setup), automation of analysis, not yet at highway speed, gigapan storage

• Implementationp– Near user ready, value added metrics aligned with current bridge rating process

16

Example of hammer sounding (green) vs. thermal IR (red):Thermal IR

Composite Willow Road bridge ThIR image

Willow Road bridge MDOT hammer sounding delamination surveyBridge deck delamination map created by thermal IR images and

output data

Optical and Thermal Image highlighting observable

subsurface defect

17

output data subsurface defect

Ultra Wide Band Imaging Radar System (UWBIRS)

Lateral translator and radar equipment. Such a 3-D system, E lE l

We did not see the results we y ,less expensive than traditional GPR, could be adapted for use on a moving vehicle.

Example Example commercial system commercial system results (Penetradar)results (Penetradar)showing showing d l i ti fd l i ti f

expected, but commercial GPR systems exist that are mapping delaminations & other bridge deterioration (hardware & as a g delaminations from delaminations from

GPR:GPR:deterioration (hardware & as a service)

18

Radar imaging of box beam interior‐ Portable Imaging Radar Mounted on a TwoPortable Imaging Radar Mounted on a Two 

Dimensional Translation Stage Parallel to Side of Salvaged Concrete Box Beam at Measurement Site at Oakland County Road Commission Site in Waterford MI

‐ Discontinuities, transitions, inhomogeneties within the structure will backscatterwithin the structure will backscatter electromagnetic energy to the radar = localized areas of enhanced radar reflectivity

‐ Used 3D Range Migration Algorithm (3D RMA) & ultrawideband, low frequency radar UWBIRS setupUWBIRS setup

‐ Inhomogeneities detected below surface –yellow areas – defects? Needs access to beam interior (cut) or beams plans to verify.  

‐ Shows promise / potential to find interior defects

‐ Could be integrated into DSS as part of condition signature score

2D cross section through the 3D radar image of the box beam viewed from the top of the box beam

2D cross section through the 3D radar image of the box beam viewed from the end of the box beam

Subsurface: Thermal IR, UWBIRS• Technology Features

– Useful metrics: % spall and delamination, detects surface and subsurface defectssurface and  subsurface defects

– Qualitative and quantitative assessment tool• LimitationsLimitations

– Collection time, camera/equipment specifics, data processing and user interpretation, cost

I l t ti• Implementation– ThIR: Near user ready, Advanced equipment, “how to deploy” manual; Radar: correlation of return with p y ;noted subsurface anomalies and further development to 3D; commercial 3‐D GPR systems available

20

Digital Image Correlation

Displacement field of cracked UHPC SpecimenScaffolding setup at Mannsiding Rd.

bridge facing t i i d ith

Laboratory set‐up

exterior girder with speckle patterns

21

Strain field of UHPC Specimen

Digital Image Correlation• Technology Features

– Can track changes in mechanical behavior over timeUseful metrics remotely captures deflection strain field– Useful metrics: remotely captures deflection, strain field and vibration (global system metric)

• Limitations– Environmental effects: error induced by wind and traffic flow, more ideally suited in current form for controlled environments

• Implementation– Not recommended for deployment without significant technology improvements (gyroscopic compensation, rigid mount), consideration of complementary technologies (laser vibrometry, LiDAR)

22

LiDAR: fixed location terrestrial surveys (MDOT)

Automated spalled area analysis

Geo-referenced LiDAR point cloud Willow Road -Section of the ortho-photo of the DEM with a color ramp applied of209 ft long. Note: Point cloud contains > 186 million points.

Willow Road bridge deck illustrating patches and a spall

DEM with a color ramp applied of the same area of the Willow Road bridge deck

23

Mobile LiDARM lik l i l f l fMost likely practical future platform: faster. Needs to be collected with bridge condition assessment in mind to get needed point densityto get needed point density

Red is higher point density – fall-offNB US-23 & ramp pfrom WB I-96 (data from 2011 MDOT survey)

24

Mobile LiDAR• Technology Features• Technology Features

– Consists of two scanners, INS, GNSS, Control unit• Portable, can be roof mounted on different vehicles, boats, rail equipmentequipment

– Data can be collected at near highway speeds• Can collect up to 500,000 points per second

LiDAR i ti t h l– LiDAR is an active technology• Data can be collected anytime traffic is light – midday, evening, nighttime.

– Technology/software still rapidly evolving– Technology/software still rapidly evolving • Survey firms already under MDOT contract – data easily available

• LimitationsLi f i h– Line of sight

– Requires high performance workstation/video card– Large files: I‐96/US23 interchange point cloud >13.7GB

25

LiDAR Summaryh l• Technology Features

– Some DOTs own equipment for non‐bridge assessment activities (familiarity with technology) or have contractactivities (familiarity with technology) or have contract access to it – just a new deployment

– Useful metrics: Deck condition (% spalled and surface condition) and Global metrics (static deflectioncondition) and Global metrics (static deflection, clearance, and post damage assessment)

• Limitations– High capital cost, speed of deployment, appropriate integration in bridge condition assessment framework

• ImplementationImplementation– Technology side is currently user ready; “how to deploy to assess bridge condition” manual needed

26

Other technologies• Satellite InSAR for bridge 

settlement– Currently limited by needed satellite ka

Riv

er

Displacement1.066 m

1.015 m

Currently limited by needed satellite imagery & reliable ground‐truth data

– May have seen ~2cm settlement at Brimley area bridge 2001‐2006 W

BR

Wai

sk

• Satellite imagery for deck surface condition– More useful for larger areas, such as  illo

w R

d

g ,longer road segments

• Airborne radar for deck surface condition (Intermap InSAR) ‐

Wi

speckle– Need good collection geometries, 

limits usefulness for bridgesan

nsid

ing

East

27

Ma

D i i S t S tD i i S t S tDecision Support System Decision Support System (DSS)(DSS)( )( )

28

Decision Support System and Result Integrationg

• Input remote sensing data can be used to create atused to create at least a bridge deck health signature

• E % spalled %• Ex: % spalled, % delaminated, roughness

• These are being integrated into an overall bridge health 

Individual bridge data along with remote sensing results Concept diagram for remote sensing datasets and their 

signature• Can be tied into 

bridge preservation 

29

p g grole in the DSS.

g pactions

29

Decision Support System and Result Integrationg

Current design:‐ Access to Bridge g

Operations tools (in the field)

‐ Access to Bridge Condition data in GIS format

‐ Access to remote sensing results –mission planning & in the field 

‐ Access to existing mapping tools

‐ Accessible via ruggedized tablets

‐ Integrated data into DSS and derived example bridge condition signature Migration to PONTIS scheme with individual bridge

attribute tablesExample integration of remote sensing data into DSS: Geo-tagged photographs from the BVRCS are available as a points layer in Bridge GIS

30

DSS remote sensing data integration(left): Screenshot from DSS of a digital elevation model (DEM) with a color-ramp and spalls outlined in red;(right): Screenshot of a high-(right): Screenshot of a high-resolution photo composite showing the same, large spall as outline on the left; note the scale bar at the bottom of each, which is 1 meter in both

Screenshot from the DSS of the hillshade DEM layer for Willow Rd bridge with spalls outlined in red;outlined in red; note that bridge joints have been removed from the spalls layer and that the vertical accuracy is so fine that patches are clearly seen

31

DSS Bridge Deck Health Signature & DSS Tool examplep

0.70.80.91.0

e

NBI R

Good

Fair

10

9

8

7

0 00.10.20.30.40.50.6

Custom

 Score

Rating

Fair

Poor

6

5

4

3

2

1

0

RS data inputs

User-changeable weights

Overall score

0.00.0 0.5 1.0

Negative Condition Indicator(e.g. % Spalled)

0

C l l ti

‐ Bridge Deck Health Signature integrates 3DOBS % spalled, IRI data + ThIR % delamination data into an overall deck health signature score with user‐changeable weights 

‐ DSS Focus Group feedback was that score inc. % spalled & % delaminated is useful to bridge management

Calculating score

DSS built from Open‐Source Technologiesg

33

llEconomic EvaluationEconomic Evaluation

34

Factors Affecting Economic Evaluation• Concept of operations

– Based costs on future, low‐cost, “drive‐by” implementationFocused on 3DOBS Thermal IR and RADAR– Focused on 3DOBS, Thermal IR, and RADAR

– Agency owned and operated or contracted service?• Time period of analysisp y

– Looked at 5‐, 10‐, and 15‐year periods (most realistic?)• Geographic scope of analysis – MI• Scale of implementation

– Percent of bridges for which remote sensing is used• Number of inspection days per year

– How many total bridges can be inspected

35

Cost Factors• Capital Costs

– Data collection system, Includes remote sensors and other dedicated hardware and software

– Data collection vehicle, installed with sensors and dedicated for inspections

• Operational Costs– Data storage and management (IT costs)– Labor

• External costs (e.g., delay, road closures, etc.)• Or Service costs – by fee• Or Service costs – by fee

36

Costing Technologies

37

Selecting Technologies• Combining Thermal IR, 3DOBS, Radar

– Up to 28 desired measurements of bridge deck surface, deck subsurface, girder surface and subsurface, global metrics (consistent, repeatable)

– Can be used in overall decision support system– Early detection of construction faults, defects, and deterioration

• Enhances routine bridge inspection• More efficient bridge scoping• Improved safety (for inspectors and traffic)Improved safety (for inspectors and traffic)

38

Cost Summary• Basic package (Thermal IR only used)

– Comes to about $530 per bridgeComes to about $530 per bridge

• More advanced system (ThIR plus 3DOBS)C t b t $590 b id– Costs about $590 per bridge

• Premier system (above plus RADAR)– Costs about $1,000 per bridge (similar to rough average cost for current inspections)

• Own and operate is more cost effective– If the technology is used for at least 70‐80 bridges per year

39

Final Evaluation of SensorFinal Evaluation of SensorFinal Evaluation of Sensor Final Evaluation of Sensor Performance Performance 

40

Scoring technologies (0‐1‐2)‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐– Meet general measurement criteria (0‐1‐2, N/A)– Commercial availability (0‐1‐2)‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐– Cost – Pre‐collection preparationC l i f l i– Complexity of analysis

– Ease of data collectionSt d ff di t ti– Stand‐off distance rating

– Traffic Disruption (no lanes closed=2, major closure=0)

41

Final Evaluation of Sensor Performance

42

Final project items

•Quarter 10 report – due June 30, 2012•Final draft report – to be submitted in July 2012•Final draft report – to be submitted in July 2012

•Closing the gap between technology and DOT use: – extending results to transportation agencies – AASHTO SCOBS meetingextending results to transportation agencies  AASHTO SCOBS meeting, 

June 2012

•Workshop on Remote Sensing for Infrastructure Management, August 24, 2012, at the ASNT NDE/NDT conference (NY)August 24, 2012, at the ASNT NDE/NDT conference (NY)

•Project ending September 30, 2012

43

Review of Project Goals and PerformanceGoal/Performance:1.Assess the potential for commercial remote sensors to be used to assess bridge condition and performance.assess bridge condition and performance.Outcome:Remote sensing has the capability to improve current bridge inspection and evaluation processes Currently some technologiesinspection and evaluation processes.  Currently, some technologies (e.g. ground‐penetrating radar and infrared thermography) are being deployed by transportation agencies, but not on a widespread basis.  The distinct advantage of using remote sensing technologies is the g g g gability to conduct non‐contact assessment without interrupting traffic.  While these remote sensing technologies do not provide a complete solution for condition assessment, they are effective in 

tif i b f h t i ti ( % ll % d l i ti )quantifying a number of characteristics (e.g. % spall, % delamination) used for condition ratings along with a visual reference and others characteristics which may be useful in future evaluations e.g. roughness patch extent)roughness, patch extent).

44

Review of Project Goals and PerformanceGoal/Performance:Goal/Performance:2.Deploy commercial sensors on in‐service bridges to assess condition.Outcome:The project team successfully deployed ## remote sensing technologies on three bridges with varying condition states in g g y gMichigan.  The scope of the deployments were limited to concrete bridges with concrete decks, but aligned well with the technologies evaluated.  The field deployment provided confidence to the p y plaboratory studies on the various technologies, but also highlighted those (3DOBS, Thermal IR, LiDAR, Gigapan, BVRCS) which appear better suited for bridge assessment and those requiring additionalbetter suited for bridge assessment and those requiring additional study (Radar, Digital Image Correlation).  The field deployment also highlighted the critical next steps for the investigation which includes progressing the key technologies to near‐highway speeds.progressing the key technologies to near highway speeds.

45

Review of Project Goals and PerformanceGoal/Performance:3.Develop decision support system (DSS) for integrating remote sensing into current bridge assessment practicessensing into current bridge assessment practices.OutcomeCurrently within the transportation community, there is no decision support system available for bridges.  A number of tools exist that have characteristics of a DSS, but none dedicated to aiding the decision making process and definitely none capable of integrating GIS referenced remote sensing data for condition assessment.  A flexible beta DSS was developed and tested with project stakeholders with great success.  The system demonstrated how remote sensing results could be collected, synthesized, and used effectively for bridge decision‐making scenarios, but also how the technologies could be integrated with current assessment methods and strategies.

46