1
Communication and Signal Processing Technologies for Intelligent Infrastructure
Arpan Pal Head of Research, Cyber-physical SystemsInnovation Lab, Kolkata
22nd November, 2012NCCCS 2012
OutlineIntelligent Infrastructure and Cyber-physical Systems
ArchitectureChallenges
Intelligent TransportationAccelerometer AnalyticsLightweight Protocols for IoT M2M Communication
Home Energy ManagementMeter Data Disaggregation
Mobile phone base WellnessAccelerometer AnalyticsMobile Camera Image Processing
RIPSAC – a generic platform for Internet-of-Things
4
Signal
Processing
Architecture
Sense
Extract
Analyze
Respond
Learn
Monitor
IntelligentInfra
@Home
@Building
@Vehicle@Utility
@Mobile
@Store
@Road
“Intelligent” (Cyber) “Infrastructure” (Physical)
APPLICATION SERVICES
BACK-END PLATFORM
INTERNET
GATEWAY
Internet-of-Things (IoT) Framework
Sense
Extract
Analyze
Respond
Communication
5
Challenges
Challenges• Reducing the cost of Communication
• Preserving the Battery power
• Extracting information from ambient-noise-corrupted sensor data
• Modeling the physics of the sensor for rich analytics
Solution Approach•Edge processing (sensor, audio, image, video) for context extraction
•Lightweight communication protocols over Internet
•Cyber-physical system model-driven system identification for analytics
7
Intelligent Transportation
M2m Cloud
Bus Tracking System
Location ( GPS), Speed, Accelerometer, Passenger Ids
Valid passenger lists, Route Info
• Can we map the Road Condition?• Can we predict Vehicle Condition?• Can we detect Bad Driving Behavior?• How to send sensor data over internet preserving
battery power and reducing network overload?
Pilot at TCS, Siruseri Campus
Business Problems
8
Vehicle Model Driven Sensor Data Analysis
CONSTANTS WE CAN MEASUREVehicle Type & Driving Behavior Road Condition MonitoringRoad Condition & Driving Behavior Car Prognosis
Road Condition & Vehicle Type Driving Behavior Analysis
Acceleration a(t) = f (H(t), v(t), R(t), D(t))
H(t)
System Identification Tools
9
Representative results : Analysis of Siruseri bus data
We compute “Road Roughness Index” for all the routes. For such evaluation, we assume basic bus model. The analysis also detects potholes /bumpers using “Jerk”.
Significant events
Spectral Analysis (10 sec window)
Computed
ISO Classification
Inference:“GOOD ROAD”
“JERK” is computed (100 sec window in the
picture) to identify isolated events
10
Validating simulation with measurementsSimulation is validated with Tata Nano car as it is forced to traverse potholes. The accelerations are
measured using SAMSUNG Galaxy Phone kept at fixed location inside the car.
Results show good match between computed pothole impact & the measured ones.. The sampling frequency is 20Hz. The deviation is amplitude is due to the approximate estimation of the impact area for pothole-tire interaction.
11
Optimized M2M communication
Enhancement & Optimization of communication cost • Enhancement of network throughput with reliable communication.• Bandwidth and energy usage optimization. • Reduction of information content.• Scalability and energy constrained are the main issues to address
12
Broadcast Based Communication
Universal Compaction using lossless source coding
Farkas, P.; Halcin, F.; , "Communication techniques for wireless sensor networks using distributed universal compaction algorithms," Signal Processing and Communication Systems, 2009. ICSPCS 2009. 3rd International Conference on , vol., no., pp.1-6, 28-30 Sept. 2009
Preservation of uplink bandwidth and sensor node power
14
Home Energy Management (HEM)
Utility
AppliancesIntelligentGateway
Smart Meter
Appliance Management
Consumption ViewOn-off Control
Social Network Integration
Consumer Home
Pilot with a Greening Company in Netherlands
• How can we identify appliances and calculate their consumption from aggregate meter data?
• How would home users share their data without being concerned about their privacy?
Business Problems
15
Objective : Disaggregate the energy meter data to get operating status of appliances connectedResearch Aspects:
– Measure active/reactive power, current/voltage waveforms using (1 Hz – 1/900 Hz) sampling rate
– Uniform smart meter data semantics by running an ontology engine - ease of data interpretation
– Pre-processing, probabilistic graphical models, temporal reasoning/data mining, pattern recognition
Non Intrusive Appliance Load Monitoring (NIALM)
Input Output
19
Sensor Collector
Mobile phone based Wellness
• How do we log exact km.s run?• How do we automatically classify the activity (Stationary, Walking,
Brisk Walking, Jogging, Spot Jogging, Running)?• How do we accurately measure the calories burnt?• How we take the instantaneous pulse rate?
Fit4Life – A wellness initiative within TCSBusiness Problems
Accelerometer
Gyroscope
Compass
Camera
20
Sensor Penetration and power consumption in Mobile Phones
0 20 40 60 80 100
Bluetooth
USB
Edge
GPRS
Wifi
3G
Camera
GPS
Accelarometer
Digital Compass
Consolidated Market Penetration
Source: Nericell: Rich Monitoring of of Road and Traffic Conditions using Mobile Smartphones, Prashant Mohan et. al., Microsoft Research, SenSys 2008, North Carolina, USA
21
Step Count using Accelerometer in iPhone
Challenges • Movement Noise Cancellation – Frequency-domain
Approach• Orientation Correction – Gyroscope and Compass based
22
Subjects Subject1 Subject2 Subject3 Subject4
Use Cases
ActualDetecte
dActual
Detected
ActualDetecte
dActual
Detected
Avg . Err%
Hand 90 76 84 83 96 91 85 70 9.9%
Shirt Pocket 90 90 86 86 93 85 88 88 2.1%
Trouser Front
Pocket90 84 85 90 95 96 89 92 4.24%
Trouser Rear
Pocket92 90 85 83 95 91 90 81 4.68%
Waist Clip 89 96 85 84 94 91 87 83 4.2%
Avg. Err% 6.4% 2.12% 4.45% 7.1% 5.024%
Next Steps • Canceling effects of Hand Movement• Activity Classification based on Step Count• Calorie-Burnt Estimation using Activity
Step Count Results – contd..
23
PPG based Pulse Measurement using Phone Camera
Challenges • Movement Noise Cancellation – Correlation with
Accelerometer• Operating in low ambient light – Advanced video pre-
processing
Subject1 Subject2 Subject3
ActualDetecte
dActual
Detected
ActualDetecte
d
68 66 66 63 85 84
2.9% 4.5% 1.1%
24
Generic Platform for IoT from TCS
Internet
End Users & Renderers
Administrators
Device Integration & Management Services
Analytics Services
Application Services
Storage
Messaging & Event Distribution Services
Ap
plic
ati
on
Serv
ices
Presentation Services
Application Support Services
OS & Device Drivers
Edge Middleware
Analytics Persistence
Application Services
Mid
dle
ware
(S
ecu
rity
/Pri
vacy
and P
roto
cols
)
Gateway
Sensors
Internet
Back-end
RIPSAC – Real-time Integrated Platform for Services & AnalytiCs
26
The Heart of Innovation – TCS Innovation LabsBangalore, India1
TCS Innovation Labs - Bangalore
Chennai, India2
TCS Innovation Labs - ChennaiTCS Innovation Labs - RetailTCS Innovation Labs - Travel & HospitalityTCS Innovation Labs - InsuranceTCS Innovation Labs - Web 2.0TCS Innovation Labs - Telecom
Cincinnati, USA3
TCS Innovation Labs - Cincinnati
Delhi, India4TCS Innovation Labs - Delhi
Hyderabad, India5
TCS Innovation Labs - HyderabadTCS Innovation Labs - CMC
Kolkata, India6
TCS Innovation Labs - Kolkata
Mumbai, India7
TCS Innovation Labs - MumbaiTCS Innovation Labs - Performance Engineering
Peterborough, UK8
TCS Innovation Labs - Peterborough
Pune, India9
TCS Innovation Labs - TRDDC - Process EngineeringTCS Innovation Labs - TRDDC - Software EngineeringTCS Innovation Labs - TRDDC - Systems ResearchTCS Innovation Labs - Engineering & Industrial Services
1 2
3
4
597
6
8
2000+
Associates in Research, Development and Asset Creation
19 Innovation Labs
27
Academic Co-Innovation Network (COIN )
Fostering joint research and innovation through a mutually beneficial alliance between TCS and academia
Academic context
Thoughts and research towards disruptive InnovationKnowledge exchange and people development
Industry-oriented Business context
innovation scalability of academia context of real-world problems
Collaborativeresearch
environment
Collaboration Measures Institution (H1 FY13)
Alliances Number of collaborations established and ongoing ISI, IIT-K, UCB, MIT, SMU (7)
Number of research alliances under consideration IISc, IITB (2)
Sabbaticals Sabbaticals from Academia Various Univs in India and Abroad (17)
Sabbaticals from TCS to Academia IIT D, Aalborg Denmark, BIT Mesra, IIT Kgp, Oxford University, Bond University (8)
RSP Research Scholars supported under RSP scheme 29 top institutes -IITs, NITs, IISc, TIFR, IIITs (100 )
IPR Papers,, Patents ISI , UC Berkley (15)
Top Related