Adaptive brake warning system for automobiles

We present an adaptive brake warning system in order to help a driver keep a safe distance from the cars in front to avoid rear-end collisions. The safe distance continuously varies according to the carpsilas velocity, driving events, and road conditions. Three sensors, an ECU reader, a camera, and a laser range finder, are used to collect the vehicular velocity, the density of cars on the road and the following distance respectively. These data are sent to a main computer to calculate the level of critical situation where the driver needs to apply brake ranging from 0 to 4 where level 0 is the safest and level 4 means the most dangerous. Two decision algorithms: adaptive threshold and fuzzy inference system are compared. The results from real driving situations indicated that both methods can achieve an accuracy of 76.98% and 65.46% respectively.