Applications of Connectivity in Automated Driving

Vehicles in the near future will be equipped with a dedicated short-range communications (DSRC) transceiver which holds great promise of significantly reducing vehicle collisions by enabling vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications. In addition, modern vehicles will be equipped with a variety of on-board sensors including GPS receivers, cameras, radars, lidars, and inertial measurement units. Using these technologies, we propose two applications to improve the driving experience and enable future advanced driver assistance systems (ADAS). First, we propose a comprehensive Kalman filter-based design that estimates the global position of the ego vehicle. Our method fuses ego vehicle’s position information obtained by the on-board GPS receiver with position information of nearby vehicles collected by the on-board ranging sensors and the messages received via the DSRC transceiver from other equipped vehicles. This process involves performing track matching using a multi-sensor multi-target track association algorithm. Second, we propose a Kalman filter that estimates the local road geometry ahead of the ego vehicle. Our method fuses on-board sensor (camera and radar) measurements with DSRC messages received from remote vehicles. This fusion system produces an accurate estimate of the road geometry on the order of 200–500 m ahead of the ego vehicle, in contrast to estimates solely derived from on-board sensors that have a typical range on the order of 100 m. At highway speeds, this increased estimation range allows the ego vehicle to reason about upcoming events 15 s before they occur in contrast to 3 s in the case of solely using on-board sensors. For both of these applications, we provide insights on the system design and present simulation and experimental results that show significant performance gains of the proposed methods in terms of localization accuracy and matching accuracy. For the road geometry estimation application in particular, experimental results show accuracy improvement by a factor of 4.8 times the current state-of-the-art camera-radar fusion methods.