Automated Highway Systems Project
Over the past 12 years CMU has been developing robotic vehicles. Currently, there is a fleet of over 10 vehicles including several cars, Army HMMWVs "Humm-vees", a minivan, and 2 buses provided by the Houston Metro system.
The original vehicle was a Chevy van that was retrofitted with computer controlled steering and brake. In the late 1980's it was capable of autonomous driving over distances of a mile at top speeds of 20 mph. And it required several racks of what were very powerful computers for that time.
More recently they used a Pontiac Minivan for a 3,000 (5000 km) drive across the USA from Pittsburgh to San Diego where it was under computer control for 98% of the way. (I think during this trip the longest stretch it went without any human intervention was about 150 miles.) During this experiment, they always had someone sitting in the driver's seat who could intervene and immediately take control of the vehicle if there were any problems. The computer that drove the vehicle was a simple lab-top PC.
One of the areas they are investigating is integrating computerized maps and satelite positional information into their system. They are currently collaborating with a company called NavTech which is the largest supplier of digitizied maps. As well, they have been using GPS information such that the vehicle can reliably locate itself. The on-board GPS (general positioning system) system processes a signal broadcast by satelite to determine its position. This allows the system to know its position at all times and determine for example if it should take an upcoming exit ramp off the highway.
While the long-term goal of this project is to develop autonomous vehicles, the more near goal is use this technology to assist the driver. They are in the process of developing a collision warning system and system that warns the driver if he or she is drifting off the road.
But, ultimately, if fully autonomous vehicles become technologically feasible and fully reliable, it is not clear what their social implications will be. So recently, 2 faculty members -- one from civil engineering and one from the EDRC (Engineering and Design Research Center) have joined this group to study the social ramafications of autonomating highways and vehicles.
Much of CMU's research in autonomous vehicles is done as part of the Autonmous Highway Systems consortium. The goal of this consortium is to develop technology to alleviate congestion and improve safety through automation. CMU's partners in this venture include General Motors, University of California, Hughes, Martin Marietta. All in all, the US dept. of transportation has budgeted over $200 million dollars for this project over the course of 7 years ($2.3 goes to CMU). This summer (Aug. '97) the consortium will put on a live demonstration of the latest technology on a section of freeway in San Diego. Among other things, CMU will participate in demonstration involving 9 fully automatic vehicles and will demonstrate obstacle avoidance, collision warning, and lane changing,
Created: Henry Schneiderman 6/97