Energy efficient robot operations by adaptive control schemes

Energy efficiency is key to achieving the Sustainable Development Goals globally. Energy consumption in the transport sector is constantly increasing, and governments are implementing policies to reduce car use by shifting the focus from roads to walking. With the rise of pedestrianization policies, autonomous mobile robots (AMRs) are becoming increasingly useful. Autonomous robotic services are being used in various fields such as traffic management, logistics, and personal mobility assistance. However, AMRs research has focused on technology development, route planning, and cost reduction, with relatively little research on how to make robots more energy efficient. As these autonomous robotic services become more popular, there is a need to discuss how to efficiently use energy. This study analyses the characteristics of the hardware required for AMRs to operate. In particular, the density of obstacles in the surrounding environment is defined as saturation for the use of Lidar, and the effectiveness of the proximal policy optimization reinforcement learning algorithm is analysed to propose an energy efficiency plan for the saturation density. In the future, a large number of robots are expected to be used, and efficient energy use of such hardware will contribute to building sustainable cities.