Sensors’ Architecture Definition for Energy Consumption Reduction in Urban Battery Electric Vehicles

Advanced Driver Assistance Systems (ADAS) have been one of the most active areas of studies in automotive sector in the last two decades. ADAS aim to support drivers by either providing warning to reduce risk exposures, or automating some of the control tasks to relieve the driver from manual driving of the vehicle. This work focuses on the definition of the sensors’ architecture for an urban battery electric vehicle (BEV) prototype, developed in the context of MOST - Centro Nazionale per la Mobilità Sostenibile, equipped with ADAS to improve its energy efficiency. An initial analysis is developed on the individual sensors, classifying them according to their characteristics, performances and actual ADAS use-case. After the analysis of the donor vehicle sensors’ setup proposed by the OEM, the benchmarking of different commercial products is carried out to highlight their best characteristics and their fit with the requirements imposed by the energy-efficient ADAS to be developed. The sensors’ chosen among the automotive-grade products are the following: Solid-State LiDAR, 4D Radar, Long-Range Radar, Short-Range Radar, Stereoscopic Camera and Monocular Camera. The final sensor kit should cover all of the ADAS functionalities to contribute to the energy management strategy, which aims to optimise comfort, battery consumption, road safety and drivability. The optimisation objective may also includes maximising battery life or, in general, a trade-off between all these objectives. The final result is a sensor suite that meets the requirements defined maximizing the different key performance indicators (KPIs) used for benchmark, such as range, Field-of-View (FoV), cost and weather robustness. The selection of the kit and its sensors is supported by hierarchical analytical process (AHP) methodology, a mathematical decision support system in which pairwise comparisons are made between the various characteristics of individual sensors. The AHP framework allows for technically quantifying the relative importance of each sensor, by providing a methodology that incorporates qualitative and quantitative notions of performance. This methodology establishes the most exhaustive sensor among those examined in the benchmark. The analysis demonstrates that the system incorporating Radar 4D yields the highest consistency index with a value of 8%.