Promoting connected and automated vehicles with cooperative sensing and control technology

An automated vehicle is a highly intelligent wheeled (maybe more than four wheels) robot that can sense its surrounding environment on the road, making navigation decisions and regulating its longitudinal and lateral motions without human intervention. Its automated system is often designed to work in an independent and autonomous way without communicating with neighboring vehicles. To achieve human -like intelligence, such a design requires a variety of high -cost sensors for reliable environment perception, a large amount of real -world training data to improve decision -making capabilities, and accurate state measurements for resilient feedback control. Recent progresses on Vehicle -to -Vehicle (V2V) communication allow automated vehicles to cyber-physically interact with each other, yielding the so-called Connected and Automated Vehicles (CAVs). CAVs could use distributed sensing, learning, and control to handle the problems faced by an autonomous system. In the layer of environment perception, distributed sensing enables CAVs to perceive their surrounding environment using not only their own sensors, but also those from neighboring vehicles. This approach can broaden the sensing range of each vehicle and achieve more accurate environmental awareness using much less expensive sensors. Distributed learning allows CAVs to share their decision rules or parameters with each other to achieve "swarm intelligence" in a more efficient way. Lastly, distributed control enables CAVs to coordinate their movements while enhancing their ability to withstand external disturbances and measurement errors. Using these distributed sensing and control techniques can further improve the safety, efficiency, and smoothness of future transportation systems.