Award Abstract #0134637
PECASE: A Framework for Driver Assistance in 2D Collision Avoidance Based on Lagrangian Systems

NSF Org:	CMMI Division of Civil, Mechanical, and Manufacturing Innovation
Initial Amendment Date:	February 1, 2002
Latest Amendment Date:	May 5, 2004
Award Number:	0134637
Award Instrument:	Standard Grant
Program Manager:	Suhada Jayasuriya CMMI Division of Civil, Mechanical, and Manufacturing Innovation ENG Directorate for Engineering
Start Date:	July 1, 2002
Expires:	June 30, 2007 (Estimated)
Awarded Amount to Date:	$375000
Investigator(s):	Joseph Gerdes gerdes@stanford.edu (Principal Investigator)
Sponsor:	Stanford University 340 Panama Street STANFORD, CA 94305 650/723-2300
NSF Program(s):	INFRAST MGMT & EXTREME EVENTS, CONTROL SYSTEMS
Field Application(s):	0308000 Industrial Technology
Program Reference Code(s):	CVIS, 1187, 1076, 1057, 1045, 1039
Program Element Code(s):	1638, 1632

ABSTRACT



Proposal Title: PECASE: A Framework for Driver Assistance in 2D Collision Avoidance Based on Lagrangian Systems

Institution: Stanford University

Recent advances such as automotive steer-by-wire systems, radar ranging sensors and differential corrections that increase the accuracy of the Global Positioning System to 20 centimeters, pave the way for active collision avoidance systems. Unlike passive restraints that seek to minimize the severity of an accident, active systems seek to move the vehicle out of harm's way, preventing an accident from occurring in the first place. Yet while the necessary components for active collision avoidance will be available in a few years, it is not clear how to integrate these into a system that smoothly assists the driver while offering guaranteed performance.

This project develops a framework for collision avoidance systems that assist the driver with lanekeeping and evading obstacles in two dimensions. The approach assigns a level of hazard to objects in the environment and applies a virtual force on the vehicle proportional to the gradient of this hazard through coordinated braking and steering. Using an energy analogy and Lagrangian dynamics, the assisted vehicle can be guaranteed to move away from obstacles while leaving the driver in control. Development of these theoretical guarantees is coupled with experimental validation of the concept on a full-scale test vehicle. The experimental equipment is also integrated into a laboratory experience in vehicle dynamics and control so that undergraduate and graduate students can literally feel the theoretical concepts of dynamics in practice. The ultimate objective of the project is to demonstrate a theoretically rigorous, yet commercially viable, system that can help reduce the approximately 37,000 traffic fatalities that occur each year, particularly the 40% due to collisions with fixed obstacles.

This project was originally funded as a CAREER award, and was converted to a Presidential Early Career Award for Engineers and Scientists (PECASE) award in May 2004.

Please report errors in award information by writing to: awardsearch@nsf.gov.