MUNICH, Germany Under the auspices of tier one Delphi Deutschland, a consortium of six companies has launched the research project 'Active Safety Car'. The project aims at developing automotive-based systems that do not only recognize obstacles ahead of the vehicle but also communicate them to other vehicles nearby.
The system will process input signals from cameras and radar sensors and interpret these data according to the needs of a relatively fast-moving vehicle. By means of car-to-car communication methods, the system will alert similar sensors systems in adjacent vehicles to watch out for such obstacles. The results are shared among the respective ad-hoc network.
This way, the researchers intend to minimize the probability that such an obstacle could be ignored due to blind areas of single vehicles' sensors. Such a system probably would improve the safety for pedestrians and cyclists, since in the sense of the project, the term 'obstacle' includes this kind of traffic participants. "We intend to provide 'intelligent safety' for all traffic participants", explained Su Birm Park from Delphi's Advanced Engineering / Safety department.
The project is motivated by a EU directive that provides to significantly reduce the number of traffic fatalities by 2010. While conventional driver assistance systems however tend to improve the safety for vehicle occupants only, the Active Safety Car project focuses on improving the safety situation for other traffic participants as well. The system is scheduled to be ready for volume production in about three years.
In terms of technology, the research will focus on real-time image processing and pattern recognition as well as on the development of a car-bound hardware platform powerful enough for this task. According to the Delhi expert, the complex software used to identify persons typically exceeds the computing power of a PC, if the results are required to be available in real-time. In addition, it will explore relevant practical aspects of car-to-car communications.
Delphi's part in the project is defining the requirements for hard- and software, system integration, and final testing. The partners of the project which is funded in part by the state government of North Rhine-Westphalia, include the Wuppertal University, Ceteq GmbH & Co KG, Riedel Communications and Volkswagen AG.
The thinking car sees a clear road ahead