Energy Conservation in Autonomous Vehicles: Challenges, Technologies, and Future Directions

Abstract

Although considered a main theme of the fourth industrial revolution, the introduction of AVs is a major hindrance to establishing a sustainable transport system. Additional driving automation levels mean increases in on-board energy consumption for sensors, computing, communication and actuator units. The energy share of AV automation is expected to be considerable, with perception, sensor fusion and real-time decision-making placing a large computational burden on the vehicle's energy, affecting the vehicle's range and emissions at the grid level. However, the performance of technology such as regenerative braking optimization, vehicle-to-everything (V2X) systems, predictive energy management, and adaptive equivalent consumption minimization strategies may achieve net savings. Due to the clear gains in powertrain efficiency and energy recovery from reinforcement learning and model predictive control systems, these control advantages are being integrated. Each of these discoveries reveals that on-road deployment of automated cars designed with energy efficiency as a co-equal engineering requirement has the potential to dramatically reduce GHG emissions from the transportation sector over a few decades.