Embedded Driver-Assistance System Using Multiple Sensors for Safe Overtaking Maneuver

Advanced driver-assistance systems have recently become one of the most active topics related to intelligent vehicles. Such assistance facilitates vehicular operation by allowing drivers increased control and an enhanced driving experience, and it even comes in the form of automated assistance for autonomous vehicle operation. This paper presents a driver-assistance system that uses a low-cost embedded digital signal processor, with the overall system installed in a commercial vehicle. Based on driving information supplied by multiple sensors, such as a real-time vision system, a vehicle-to-vehicle communication system, and in-vehicle sensors, the proposed system can facilitate decision making and the performing of driving tasks while executing overtaking maneuvers. This paper developed a data fusion stage based on a collision warning algorithm in which the overtaken vehicle and other vehicles in the neighboring lane are accounted for to avoid collisions. The system employs fuzzy control in the steering and speed automation to emulate the driving tasks performed by humans. The applicability of the proposed system was examined in a real-road environment, and a set of experimental results demonstrate the feasibility of using the involved coordination strategies while conducting various driving maneuvers.