Dtof LIDAR System Using Addressable Multi-Channel VCSEL Transmitter, 128x80 SPAD Sensor, and ML-Based Object Detection for Adaptive Beam-Steering

Light detection and ranging (LiDAR) technology is increasingly being deployed in vehicles for advanced driving assistance system (ADAS) and is a potential critical technology for autonomous driving in the future. Currently, hybrid solid-state LiDARs based on micro-electro- mechanical system (MEMS) or rotating mirrors are the most mature solutions [1]. Unfortunately, these hybrid LiDARs still exhibit high risk of damage due to their limited mechanical reliability in harsh environments across many years of operation. Furthermore, these solutions require highly accurate calibration due to the complex optical alignment, resulting in higher production cost. Compared with these mechanical solutions, solid-state flash LiDARs without any moving parts are easier to mass-produce, more robust, less costly, and exhibit smaller form-factor, allowing flash-LiDAR to be embedded in multiple places around the automobile. Nevertheless, the detection distance of a single-flash LiDAR is short due to the smaller optical power limited by the laser eye-safety requirement, VCSEL thermal heating, and long-term VCSEL reliability [2]. Furthermore, for some applications, the distance of only one or two nearest objects needs to be accurately measured, and therefore empty parts of the field-of-view (FOV) are of limited interest. Unfortunately, single-flash LiDARs capture the point cloud of the entire FOV within each frame, resulting in high power consumption.