Safety Co-Pilot: A System for Autonomous Vehicle to Make Decision Safer and Smarter

The Autonomous Vehicle (AV) has become a hot research spot recently. Following the trend of auto driving (AD) development, making AV safer is evolving as a new direction of research. The decision system in AD can be regarded as the brain of the vehicle, which ensures safety and comfort in driving behaviors. However, technologies of decision systems, including algorithms and hardware, remain incomplete and could affect the safety of the AV. Hence, it is necessary to focus on the decision safety of AD systems before AVs are in widespread application. In this paper, the existing techniques for vehicle decision system safety real-time assessment and response are briefly reviewed first. Based on the current research, we propose the “safety co-pilot (SCP)” intelligent decision safety system to achieve better supervision of AD systems. To comply with the development tendency towards the V2X of AVs, the cooperation of the system with online cloud learning is fully expounded, and the outlook and challenges of this safety system are deeply analyzed. The proposed SCP system is supposed to not only increase the real-time driving safety of AV but through communicate with the decision system constantly to further improve the intelligence and reliability of the AD system.