Towards cooperative, predictive driver assistance

In this work, we describe research directed towards cooperative, predictive driver assistance. We describe predictive driver assistance systems, which aim to make longer-term predictions about the on-road environment, supporting positive HMI with the driver. Given the rapid development of advanced sensing, computational, and algorithmic technologies, intelligent vehicles are now approaching an era in which the vehicle can maintains a full panoramic awareness. Thus, the instrumented vehicle will be available to help the driver navigate the on-road environment, when the driver requests it. As predictive driver assistance systems are developed, it will be shown that cooperative integration, across vehicles, will improve the performance and overall on-road safety. We divide the scope of cooperation systems into four main areas: in-vehicle systems cooperation, vehicle-driver cooperation, cooperation across vehicles (V2V), and finally cooperation with vehicles and infrastructure (V2I). Each level of cooperation will quantitatively improve on-road safety performance, and allow for increased sophistication and prediction.