Investigation on Control Methods and Development of

Intelligent Vehicle Controller for Automated Highway

Systems

P.Suresh

ME11D045

Guide

Dr. P. V. Manivannan

Precision Engineering and Instrumentation Laboratory

Department of Mechanical Engineering

Indian Institute of Technology Madras

Chennai – 600036

1/15

Contents

Introduction

Intelligent Vehicles

Objectives of Research

Success Stories

References

2/15

Contents

Introduction

Challenges in Road Transportation:

Traffic Congestion

High accident rate

High Accident mortality

Increase in Vehicle Population

Government initiatives to tackle surface transportation problems:

Conversion of 2 lane highways to 4 lane highways

Building high speed expressways

Golden Quadrilateral project

Strict enforcement of traffic regulations

These Initiatives have not been able to solve the problems.

3/15

Introduction

The Single Most Important Factor for Road Accidents is:

Human Error

Drivers tend to commit errors when driving at very high speeds due to the

following reasons:

Fatigue

Poor Visibility

Deteriorated Road Conditions

Adverse Weather

Very small available response times

Intelligent vehicles and Automated Highway Systems (AHS) will help in

reducing the number of accidents and thus minimise loss of life and property

resulting from accidents.

4/15

Introduction

History of AHS & Intelligent Vehicles 1939 – New York World Fair – GM Futurama

first formal introduction to the idea of AHS and autonomous vehicles

1953 – scale model of AHS developed by GM and Radio Corporation of America

1958 – GM tests full size passenger car with in-built guidance system

1960s – Development of road centralised control system (Dr. Valdamir Zworykin from RCA).

Circuits buried in the road to magnetically sense vehicle speed and location.

1980s – a vision-guided Mercedes-Benz robot van, designed by Ernst

Dickmanns, Bundeswehr University Munich

DARPA-funded Autonomous Land Vehicle (ALV) in the United States

1987-1995- EUREKA Prometheus Project on autonomous vehicles

2000- US Army funded DARPA DEMO I,II and III projects for autonomous vehicles over difficult road terrain.

2000-AHSRA Demo 2000 (Japan) - 38 cars, buses and trucks illustrated the ideal system for reducing road traffic

accidents using driver information and control assist systems. The automation system made use of magnetic

sensors on the road.

2001--the Carnegie Mellon University Navlab project-semi-autonomous car

2001-2003-Chauffeur II –development of truck platooning by DaimlerChrysler, Renault, IVECO and Fiat

2009- SARTRE project-UK, Spain & Sweden- Platooning system with lead vehicle controlled by a professional driver

2010 -VisLab ran VIAC, the VisLab Intercontinental Autonomous challenge, 13000 KM test run of autonomous vehicles.

2011- First successful trial of SARTRE project was in Jan2011 at Volvo Test Track in Sweden-single car was slaved

behind a rigid truck.

1994- twin robot vehicles Vamp and Vita-2 of Daimler-Benz drove semi autonomously

1995-the Carnegie Mellon University Navlab project-semi-autonomous car

1996-Alberto Broggi of the University of Parma launched the ARGO Project – lane following

achieved

1997-DEMO 97- Fleet of over 20 vehicles guided on a 7 mile stretch of Interstate 15 Highway north of

San Diego, USA

5/15

Introduction

Intelligent Vehicles (Driverless Cars)

Vehicles equipped with an autopilot system --- capable of driving without

input from a human driver.

Advantages of intelligent / autonomous vehicles are:

Relief from driving and navigating task

Accident prevention

Increased roadway capacity

Traffic congestion reduction

Increased safety due to elimination of driver error

6/15

Intelligent vehicles

Intelligent Vehicle Sub-systems

Master controller

Steering controller Braking controller

Obstacle

avoidance

controller

Engine

controller

(ECU)

Transmission

controller Lane following

controller

7/15

Intelligent vehicles

Fig. 1 Vehicle sub-systems

Sensors

8/15

Intelligent vehicles

Perception is an important aspect of any intelligent or autonomous

system.

Some of the sensors used in the Intelligent Vehicles are:

GPS (Global Positioning System)-

• provides the absolute location and direction of the vehicle on the

road

• data is received from satellites orbiting the earth

Optical Camera - Eye of the vehicle. Provides it vision capability

Infra red camera - Provides night vision capability

Radar – Measurement of distances (vehicles -vehicle and Vehicle – obstacles)

Sensors

9/15

Laser Scanner – Known as LIDAR (Light Detection And Ranging) Sensor.

• Most popular sensor, used in most vehicles.

• Very expensive (a typical unit costs more than 3 lakh

Rupees)

• Provides 2D and 3D data of the vehicle surrounding

environment

Odometry – measurement of changes in position, velocity and acceleration

using sensors located in moving parts.

Inertial Measurement Systems – Accelerometers and Gyroscopes

• Measures relative movement of robot in linear or angular

direction.

Compass – Determines vehicle direction with respect to earth’s poles.

Intelligent vehicles

Sensors

10/15

Sensor fusion – Mechanism or algorithm that combines the data from

different sensors into one perception of the environment

Microwave

Radar

LIDAR

Infrared camera

Sensor

Fusion

Mechanism/

Algorithm

Driving

software

Perception

Optical camera

GPS

Gyroscope

Wheel encoder

Compass

Vehicle

mechanics

Intelligent vehicles

Fig. 2 Vehicle sensor fusion mapping

Intelligent Vehicle Sensor Location

(Figure from Robotland Blog article “Pickup an autonomous taxi cab in Berlin with iPad” dated October 18 th,2011) 11/15

Intelligent vehicles

Autonomous Passat

by

Volkswagen, Germany

Fig. 3 Autonomous Passat

Vehicle Architecture

12/15

Sensor

Sensor

Sensor

Sensor Fusion

Mechanism/

Algorithm

|

|

|

Driver

Behavior

Model

Perception

A general architecture of an intelligent vehicle.

Path Planner

Navigator

Driver / Pilot

Engine

controller

Route map &

destination information

Brake

Controller Transmission

Controller

Steering

controller

Driving

software

Physical

Vehicle

layer

Intelligent vehicles

Fig. 4 Vehicle Architecture

Replace the Human Driver with an Equivalent Intelligent

System:

Develop a Supervisory / Master controller that will co-ordinate the

functioning of the various sub-system controllers of the intelligent

vehicle.

• Development of Vehicle and Driver behavior Models (Finite State

Machine – FSM)

• Development of control algorithms : Fuzzy, Neural and Genetic

Algorithms

• Simulation and validation of developed control algorithms

• Development of Embedded controller

• Testing with scaled down vehicle model

Objectives of Research

13/15

Objectives

14/15

Actual Implementation – Success Stories

Success Stories

Google’s Autonomous Car project

• Fleet of robotic Toyota Priuses have covered more than 1,90,000

miles (3 lakh kilometers) in all types of road conditions with minimal

human input.

University of Berlin’s “MadeinGermany” VW Passat

• Autonomous taxis in Berlin

VisLab (Italy) – VisLab Intercontinental Autonomous Challenge

(VIAC)

• Four driverless vehicles -15,000 KM trip - Parma in Italy to Shanghai

China - July 26, 2010 - October 28, 2010 with virtually no driver

intervention.

1. Massimo Bertozzi, Alberto Broggi, Alessandra Fascioli, ‘Vision-based intelligent vehicles: State

of the art and perspectives’, Robotics and Autonomous Systems 32 (2000) 1 – 16,

01/02/1999.

2. Nikhi M Chakravarthy and Lawrence B Holder Jr, ‘Intelligent Cars’, Intelligent Environments

Presentation, Computer Science and Engineering Department, UTA, Spring 2003

3. Jameson M Wetmore, ‘Driving the Dream: The History and Motivations Behind 60 Years of

Automated Highway Systems in America’ , Automotive History Review, Summer 2003, pp. 4-

19.

4. Sadayuki Tsugawa, ‘A History of Automated Highway Systems in Japan and Future Issues ‘,

Proceedings of the 2008 IEEE International Conference on Vehicular Electronics and Safety,

Columbus, OH, USA. September 22-24, 2008

5. Ola Ringdahl, ‘Techniques and Algorithms for Autonomous Vehicles in Forest Environment’,

Dept. of Computing Sciences, Umea University, Sweden, ISSN-0348-0542, ISBN-978-91-

7264-373-4.

6. John Leonard, Jonathan How, Seth Teller, David Barrett, Chris Sanders,‘DARPA Urban

Challenge: Team MIT Development Plan’, October 27, 2006

7. Robotland blog,’Pick up an autonomous taxi in Berlin with iPad’,October 18,2011

8. Erico Guizzo,’How Google’s Self-Driving Car Works’,IEEE Spectrum, October 18,2011

http://spectrum.ieee.org/automaton/green-tech/advanced-cars/how-google-self-driving-car-

works

References

15/15

References

THANK YOU