Radisys is the computer of the intelligent highway.
A consortium funded by the government, GM, and others will be demonstrating an automated group of cars, trucks, and buses on a highway near San Diego Aug. 7-10. The computer system in each car contains 2 Pentium-166's from Radisys. The estimated system deployment timeframe is 2005-2010.
Radisys' role in the project is described in the company magazine at radisys.com (Adobe Acrobat format).
The intelligent highway works by cramming cars closer together. The goal of the National Automated Highway System Consortium is to double or even triple the density of cars on existing highways. The platooned cars (convoy) could be spaced as close together as 2 meters! The computers control the accelerator, brakes, and steering and the cars talk to each other. The San Diego demo will use 7.5 miles of carpool lanes on I-15 and automated and manually operated vehicles will share the road.
Several of the subsystems have already been deployed. For instance, "adaptive cruise control" is already used in Japan and the car automatically adjusts your cruise control speed based on the distance between your car and the car in front of you. Some cars also can detect when you drift into the next lane. Anti-lock brakes provide the brake actuator. The target cost of the complete system is $800-1000, about the same as airbags a few years ago (today?).
There's an interview with the project director at itsonline.com that describes the demo and is partially reproduced below. I normally don't advocate copyright infringement, but my tax dollars paid for this project:
1. Free-agent, multi-platform scenario. That will combine two transit buses from Houston Metro and several passenger vehicles. These vehicles will operate as free agents, but they do communicate with each other. As we discussed before, we make a distinction between free agent operation and platoon operation. This scenario will show vehicle following and totally autonomous operation. The buses and cars will have computer-controlled steering, braking and throttle. They will run down the 7 1/2 miles according to a pre-scripted scenario, where the buses do automatic lane changes and pass the cars, and the cars do automated lane changes and pass the buses. Each detects and avoids obstacles and communicates information about itself to the other vehicles involved. Computer vision technology will be used for lateral or lane control.
2. Platooning Scenario. This will involve 8 passenger cars that operate in a closely placed platoon -- less than 10 meters apart. They will accelerate, decelerate, show split into two sub-platoons to allow other vehicles to enter, and then rejoin as one platoon. They will also show a coordinated stop, as if an obstacle were encountered.
3. Heavy Truck Scenario. We'll have a passenger car and one or two full-sized class 8 commercial vehicles that will show adaptive cruise control, collision warning, and side obstacle detection. Some passing maneuvers will be demonstrated as well.
4. Evolutionary Scenario. Using passenger cars, we will show the stages in evolution, starting with adaptive cruise control and evolving to collision warning, collision avoidance, and full automated vehicle control. This involves several demonstrations executed consecutively on the 7.5 mile stretch of I-15.
5. Transition of Control Scenario. This scenario shows the way a vehicle might move from a rural setting, where it was depending mostly on its own resources, to an urban setting where it communicates with the highway. These vehicles will start with computer vision for lane sensing, as if they were in a rural area, then transition to magnetic nails (markers) in the freeway, as if they were in an urban setting.
6. Alternative Technology Scenario. Instead of using computer vision or magnetic nails in the roadway, this scenario will start off with computer vision and then switch to a radar-reflective tape installed on the roadway. Radar reflective tape offers the potential of using the same sensor for collision warning as for lateral control. |
