Full beam on driver assistance systems
Advanced driver assistance systems (ADAS) will make life a whole lot easier and safer for the driver and of course, other road users. At present, they are only usable in the relatively simple driving situations. On a motorway, for example, processing can be focused on a few well-defined detected objects. However, in dense traffic situations in urban areas, the situation is littered with traffic lights, signs, tight bends and bikes weaving in and out, to mention just a few rogue unexpected elements. Even in motorway traffic, one major drawback of ADAS is the phenomenon of cut in, where vehicles are placed in the detection beam close to the car. Unfortunately, the width of the sensor beam does not cover the area in front of the car. Another drawback is the inability to deal with fixed objects. Looking to the future, the EU-funded project Carsense aimed to develop existing ADAS like Adaptive cruise control (ACC) more effective and applicable in normal everyday driving conditions. Project researchers concentrated on limitations of available systems, in particular sensing functions, to improve their coping ability in difficult scenarios. Carsense developed sensor systems in a flexible architecture to deal with complex urban situations at low speed. They came up with an ingenious method to calibrate sensors and test their algorithms. Data from various sensors is collected simultaneously and recorded with driving sequences scripts. A project partner, Laboratoire Central des Ponts et Chaussées, used its own tracks to undergo the testing. A second set of trials provided the ultimate test for sensor fusion and the performance of the new sensors. Mulitple sensor information improves detection reliability. Carsense sensors can also pick up fixed objects and have a wider scope. Applications include collision mitigation where, for example, brakes cut in to lessen the impact of a crash. The ultimate goal of complete collision avoidance is also a distinct possibility using the new integrated sensor systems.
