on open source mobile sensing - pdfs.semanticscholar.org · open source libraries for mobile...

Post on 21-Aug-2020

5 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

On Open Source Mobile

Sensing

Dmitry Namiot Lomonosov Moscow State University

dnamiot@gmail.com

Manfred Sneps-Sneppe ZNIIS, M2M Competence Center

manfreds.sneps@gmail.com

ruSMART 2014

Phone as a sensor model.

• Smart phones as an ideal platform for collecting

and processing context-related data.

• Computational social science, crowdsensing

• An an attempt to describe and categorize existing

open source libraries for mobile sensing,

• Describe architecture and design patterns

• Discover directions for the future development.

About

Contents

Introduction

On challenges for mobile phone sensing

Open source libraries for mobile phone sensing

The model and patterns

Conclusion

Introduction

• Rich sensing capabilities for smart phones

• Collecting data about people’s social behavior

(computational social science – e.g., Reality

Mining)

• Crowd-sensing for business-related tasks (e.g.

OpenSignal)

• Balance between energy efficiency, data

collection, storage, and transmission procedures

Challenges

• Batteries as a major challenge in achieving

social sensing

• Sensors power consumption: GPS vs.

accelerometer

• Context-aware data collecting. E.g. reuse

location for phone on the table, SD card vs.

cloud storage, etc.

• High level of diversification in mobile sensors

External collections

Open Source Frameworks

• AWARE framework: client + server

Open Source Frameworks

• FUNF framework

Open Source Frameworks

• Open Data Kit

Challenges for Open Source

Frameworks.

• Context-aware data collecting. How to

reduce measurements and data

transmission

• A flexible data management. SD-card vs.

Cloud

• Portable (common) data formats

• Built-in data processing

API vs. DPI

• Traditionally: mobile OS presents API for

built-in sensors

• APIs used by mobile applications

• The standard approach for crowd-sensing

is to split data collecting and data

processing

• So, we have to switch to DPI – Data

Programming Interfaces

Conclusion

• A survey of the Open Source tools

for mobile sensing.

• Existing projects

• Directions for the future research

• The prediction: we will see mobile sensing as a part

of mobile OS

• An existing example: iBeacons в iOS

About us

International team: Russia - Latvia (Moscow –

Riga – Ventspils). Big history of developing

innovative telecom and software services,

international contests awards

Research areas are:

open API for telecom,

web access for telecom data,

Smart Cities,

M2M applications, context-aware computing.

top related