Defense and Civilian Applications of the ALVINN Robot Driving System
Abstract
Perhaps the most successful development of the ARPA Unmanned Ground Vehicle (UGV) program is the ALVINN autonomous driving system. ALVINN is a neural network system that locates the road in images from a video camera mounted on a UGV and steers the vehicle to follow it. ALVINN has been demonstrated on several HMM WV test vehicles driving at speeds of up to 70 mph, and for distances of over 90 miles without human intervention. ALVINN was originally designed as part of an unmanned vehicle for the modern battlefield, performing reconnaissance, surveillance and nuclear, biological and chemical (NBC) detection missions. However we are currently adapting the system for civilian use, as part of the Intelligent Vehicle Highway System (IVHS) initiative. This paper describes the ALVINN system, and its potential applications to both defense and civilian applications. It also describes the the technical issues involved in converting the system from military to civilian operation, and the steps we are taking to address them.
BibTeX
@conference{Pomerleau-1994-13789,author = {Dean Pomerleau},
title = {Defense and Civilian Applications of the ALVINN Robot Driving System},
booktitle = {Proceedings of Government Microcircuit Applications Conference (GOMAC '94)},
year = {1994},
month = {November},
pages = {358-362},
}