Benchmarking LiDAR Sensors for Development and Evaluation of Automotive Perception

Environment perception and representation are some of the most critical tasks in automated driving. To meet the stringent needs of safety standards such as ISO 26262 there is a need for efficient quantitative evaluation of the perceived information. However, to use typical methods of evaluation, such as comparing using annotated data, is not scalable due to the manual effort involved. There is thus a need to automate the process of data annotation. This paper focuses on the LiDAR sensor and aims to identify the limitations of the sensor and provides a methodology to generate annotated data of a measurable quality. The limitations with the sensor are analysed in a Systematic Literature Review on available academic texts and refined by unstructured interviews with experts. The main contributions are 1) the SLR with related interviews to identify LiDAR sensor limitations and 2) the associated methodology which allows us to generate world representations.