The founder of the world's top LiDAR maker says there's one problem with Elon Musk's approach to self-driving (www.businessinsider.com)

Steven Qiu, founder and chief scientist of LiDAR maker RoboSense, told Business Insider that multi-sensor stacks (including LiDAR) are essential for safe Level 3/4 autonomous driving, directly challenging Elon Musk’s long-standing claim that LiDAR is “expensive and unnecessary.” LiDAR—which emits laser pulses to measure distance—fills perception gaps that cameras struggle with, and Qiu argued vision-only systems cannot reliably handle many corner cases. RoboSense, which led passenger-car LiDAR market share in 2024 per Yole Group, notes LiDAR costs have plummeted from roughly $70,000 per vehicle to the low hundreds while performance has risen, making sensor fusion more feasible commercially. The technical implications matter for the AI/ML community: sensor fusion and robust perception models must integrate LiDAR point clouds with camera imagery to resolve ambiguity (e.g., distinguishing a white stopped car from a cloud, detecting a dark vehicle entering a tunnel, or overcoming sun glare and reflections). Industry peers like Ford and China’s Li Auto also back LiDAR for real-world traffic complexity, highlighting that dataset diversity, multimodal architectures, calibration, and edge inference will be central to advancing beyond Level 2 systems such as Tesla’s FSD. In short, the debate shifts from whether to use LiDAR to how best to fuse sensors and train models to handle rare, safety-critical scenarios.