Intelligent Transportation Systems

(ITS)

Lelitha Devi Vanajakshi

Professor
Dept. of Civil Engg.

[email protected]
Intelligent Transportation Systems (ITS)

• Application of well-established technologies in

communications, control, electronics, and
computer hardware and software to improve
surface transportation system performance.
(Source: Perspectives on ITS, J. M. Sussman)
 How it work
1. Need to sense the presence and identity of vehicles in
real-time on the infrastructure through automated
sensors;

2. Need to communicate (i.e., transmit) large amounts of

information more cheaply and reliably;

3. Need to process large amounts of information in an

accurate and reliable manner; and

4. Need to inform users in real-time in order to achieve better

transportation network operations.
 Components
• Sensors

• Communication

• Computing/Analytics/Modeling

• Actuators/Solutions/Applications
 Placement
System components
✓ Vehicle
✓ Way
✓ Control
✓ Human

Sensors in each of these components

5
ITS International
 Sensor Classification
• Roadway sensors/ Location based sensors
– Macroscopic data
– Eulerian
– Eg. Video

• In-vehicle sensors/spatial sensors

– Microscopic data - Tracking
– Lagrangian – trajectory
– Eg. GPS

• Combination
– Re-identification
– Eg. RFID

• Crowd Sourcing
– User/Driver based data
– Eg. social media
 Location Based Sensors
• More popular, Easy to manage

• Invasive/intrusive
– Sensor placed on or in the road
– Pneumatic, Magnetic, Piezoelectric, Inductive loop

• Non invasive
– Non-intrusive techniques – roadside or above
– Radar, Infrared, Video etc.
 Issues – Location Based
• Miss spatial variation.

• Invasive ones - Repair due to wear and tear. Repair is

difficult and affects traffic.

• Non-invasive ones need unobstructed view of traffic -

Occlusion

• Performance vary with sensor type, mounting height,

location, vehicle mix, road configuration, viewing angle
etc.

• Aesthetic issues, Maintenance affect traffic

 Tracking /Re-identification
• AVI
• GPS
• Cell phones
• OBD
• Connected vehicle sensors
• Bluetooth/Wi-Fi
• RFID
 Issues – Spatial

• Privacy issues

• Need participation

• Leads to sample size limitation, and bias

• Cost and safety issues

 Indian Context
• Existing location based sensors and solutions
developed for lane based and homogeneous

• Indigenous solutions specifically taking into

account Indian conditions
– Heterogeneity
– Lack of lane discipline
– Variability
– Randomness
 Development of New Sensors

• Inductive Loop Detector

• Bluetooth/Wi-Fi Sensor
• Image processing
• Crowd sourcing/Social media

Social Gears
 New Inductive Loop Detector

➢ New shape and configuration for Indian traffic

INDUCTIVE LOOP SENSOR DEVICE, Patent number: 341092.

 The New Inductive Loop Sensor

Large vehicle (e.g., bus)

Relative change
in inductance

Small vehicle (e.g., bicycle)

Position of the object

Loop-A Loop-B Loop-C

Results from the New Single Loop Detector

CAR

BUS

Output signal observed for different types of vehicles

Multiple Loop Detector System
Bluetooth Sensor

Sample Output

17
Working Principle of a WMS

Scanning Module
WiFi MAC Scanner

Computation
Module

Storage Module

Communication
Module
 Image Processing

– Processing for classified

count, speed, occupancy,
queue length at signals, CAR

spacing
– Congestion quantification
– Red-light violation
– Deep learning
 Steps
(a) Sample Frame (b) Region of Interest

– Camera Calibration
– Background Subtraction
(Foreground Extraction)
(c) Output of Background Subtraction (d) Image with all the edges

– Shadow Elimination
– Detection
– Tracking
(e) Outer edges from foreground mask (f) Image with the inner edges

(g) AND operation on the derivatives (h) Final extracted Foreground

 Tracking for Speed

(a) Sample Frame (b) Frame with detected corner points

(c) Before applying thresholds (d) After applying thresholds

 Video Analytics for RLV

(a) (b) (c) (d) (e)

(a) selected frame, (b) selected region of interest, (c) output of the background subtraction step,
(d) output of shadow detection step, (e) final foreground mask obtained

RLV
Data from Social Media

12
 Data from Social Media
Data Collection
• Twitter
• Collecting instances of
textual data

Text Processing
• Pre-filtering
• Text-mining techniques

Feature Extraction Feature Extraction

Machine Learning based

Lexicon based Approach

• Labelling tweets
• Labelling tweets
• Primary and secondary
• TF-IDF
keywords

Approach
Tweet Classification Tweet Classification
• Word-patterns classifier • Trained classifier for
with OR & AND operations incident type detection
Traffic Information
• Traffic information in
desired classes
24
Crowd Sourcing – Mobile Apps
• Bus drivers, bus passengers, and
travelers *

• GPS logger module - The latitude and

longitude information is collected from cell
phones

• Connects to Amazon Web Services

database service, and stores data

• Can be bundled with other applications

* In collaboration with TCS Innovation Labs, Chennai

 Communication
Communication
✓ Vehicle to Vehicle
✓ Vehicle to Network
✓ Vehicle to Infrastructure
✓ Vehicle to Pedestrian
✓ Vehicle to Devices

GPRS
DSRC
Cabled connection
Wireless communication
26
 Analytics/Modeling
• Visualizations
• Quality Control – Outlier removal,
imputation, smoothing
• Statistical analysis
• Empirical/Data driven Models
• Mathematical Modeling
• Big data analytics
 Functional areas
• Advanced Traffic Management Systems (ATMS),
ATC
• Advanced Traveler Information Systems (ATIS),
Dynamic route guidance
• Advanced Vehicle Control Systems (AVCS),
Collision warning/avoidance system
• Commercial Vehicle Operations (CVO),
Truck – vehicle health monitoring
• Advanced Public Transportation Systems (APTS),
Bus arrival prediction
• Advanced Rural Transportation Systems (ARTS)
Advanced dispatching and vehicle-based response systems
(e.g., on emergency medical services)
Applications
 Applications Developed
• ATIS
– Advanced traveler information system
– Congestion prediction system
– Route planner
– Departure time planner
• ATMS
– Isolated, Coordinated and Area traffic control
– Bus priority system
– Automated tolling solution
– Automated red light violation detection
• APTS
– Real time bus tracking system
– Bus arrival time prediction system
– Dynamic trip planner
 Facilities @ IITM
• ITS Laboratory

• Instrumented test bed

• Simulator and VR Laboratory

• Connected Vehicles
Futuristic
 Autonomous vehicles

https://www.vox.com/2016/7/5/12077002/self-driving-cars-improve-cities
 GLOSA
• “Green Light Optimal Speed Advisory”
• Uses information on traffic light timings from
a roadside unit to display to the driver the
best speed to travel to get a green light.
 Autonomous Intersection Management

• Reservation Based

www.cs.utexas.edu
 Organisational issues
• Highly multi-disciplinary - requires unprecedented levels of
cooperation and coordination between agencies to take it to field and
show real world success.

• Difficult transition - From conventional Civil Engineering technologies

of structures, materials, geotechnical engineering and project
management, to electronics, information systems, communications and
sensors.

• Substantial leadership required - to implement such projects as an

integrator of transportation, communications and intermodalism