In thess years, vehicle LiDAR occupies the central stage in the technology roadmap of level-3 to level-5 autonomous driving. Due to its expected enormous market scale, resources and efforts from industry and academic to develop a cost-affordable and vehicle-spec LiDAR. The LiDAR modules in current market mostly use one-axis or 2-axis mechanical scanning mirrors, either macroscopic or microscopic, to proceed 3-D imaging. The moving parts in these LiDARs may result in safety risks with vibrant condition and/or reliability issues in long term so have to be eliminated eventually. In our group, an intelligent vehicle LiDAR without any moving part is to be developed. By using the high-sensitivity and low-cost CMOS single-photon avalanche diode (SPAD) array and the high-power and low-divergence-angle photonic-crystal surface-emitting laser (PCSEL) array, we shall investigate the feasibility of a sequential flash scheme by synchronizing the PCSEL emission and the SPADs detection. Beside, the obtained 3-D point cloud will be used to do object detection with the integrated artificial intelligence (AI) chip. The AI chip with accelerator will be developed to accomplish real-time object detection for on-road scenes. Through this three-year project, we shall develop the core technologies and know-hows for an intelligent vehicle LiDAR without moving parts, which will pave the way for autonomous driving in the near future.
