Elon Musk is making a strong case against the use of lidar technology in autonomous driving systems. In a recent statement, the CEO of Tesla argued that not only is lidar unnecessary for achieving self-driving capabilities, but it can actually make self-driving cars more dangerous.
Musk’s comments come in response to Uber CEO Dara Khosrowshahi’s support for the use of lidar and radar in autonomous vehicles. Khosrowshahi believes that outfitting cars with sensors like lidar is essential for achieving superhuman levels of safety in self-driving cars. He pointed out that many of Uber’s partners are using a combination of cameras, radar, and lidar in their autonomous systems.
However, Musk disagrees with this approach. He argues that the use of lidar and radar can actually reduce safety by creating sensor ambiguity. When sensors like lidar and radar disagree with cameras, it can lead to increased risk rather than decreased safety. Musk cited Tesla’s own experience, stating that the company saw an improvement in safety once they switched to a vision-only approach for their autonomous systems.
Tesla has been able to launch a successful Robotaxi pilot in Austin and the Bay Area using their vision-based autonomous technology. Their Full Self-Driving (FSD) system is also rapidly advancing and is becoming increasingly better at driving than humans.
Musk’s stance on lidar technology has sparked a debate in the autonomous driving industry. While some companies like Waymo continue to rely on lidar for their self-driving systems, Tesla’s success with vision-based technology is challenging the conventional wisdom in the field. Musk’s comments highlight the importance of carefully considering the role of different sensors in autonomous driving systems and the potential risks associated with their use.