CARS THAT automatically avoid other vehicles, pedestrians and obstacles could be a step closer thanks to a Society of Motor Manufacturers and Traders (SMMT) Foresight Vehicle project called AutoTaxi.

The aim is to develop safety-critical sensors and systems to process and interpret critical information about a car's local environment - for example, the path of the road ahead and the presence and movement of vehicles and other obstacles on it. It could pave the way to fully autonomous cars of the future.

Project manager Alastair Buchanan of TRW Conekt explained: "Developing cars that do things which are second nature to the average driver is a hugely complex process.

"The vehicle has to be able to sense and interpret the vast array of information in all driving conditions. AutoTaxi will help identify the best combination of sensor technologies that may take this concept a step closer to reality."

Sensor technologies are already bringing real-world benefits to drivers of new cars. Ultrasonic parking sensors are now common, while radar in Active Cruise Control (ACC) sets speed between vehicles on high-speed roads.

AutoTaxi took sensors used in automotive applications, and applied different combinations to ULTra vehicles operating on ATS's Cardiff test track.

Starting with simple assessments on a figure of eight track, the tests investigated more complex scenarios, to simulate real world driving conditions such as vehicles approaching from side roads and parking.

Phase one of the project linked TRW's video lane guidance and ACC radar with lidar, which uses light to measure distance and speed, as well as ultrasonic sensors. These were fine-tuned with the addition of stereo video sensors and an LED rangefinder.

Trials in November revealed potential particularly for the LED rangefinder system. The data will help develop algorithms to interpret the information with results due this summer.