GM Exec Says Ditching Radar for Self-Driving 'Ignores Physics'
Chief Product Officer Sterling Anderson labels camera-only approaches a 'weird religious stance' ahead of 2028 Super Cruise launch.
General Motors Chief Product Officer Sterling Anderson has criticized the industry trend toward camera-only autonomous driving systems, arguing that removing radar from the sensor suite is a decision that ignores physical limitations.
What Happened
In an interview with The Verge, Anderson described the move to scale back to vision-only systems as a "weird religious stance that ignores physics." He emphasized that radar is inexpensive and provides critical capabilities, such as measuring instantaneous velocity and seeing through dense fog, which cameras cannot replicate. Anderson, who previously led Tesla's Autopilot program and co-founded self-driving startup Aurora, stated that a multisensor suite provides "obvious safety advantages that physics simply demands."
The comments come as GM prepares to launch an enhanced version of its Super Cruise system in the 2028 Escalade IQ. This upcoming feature is designed for highway use and will utilize a combination of radar, lidar, and cameras. Unlike Tesla’s approach, GM intends to maintain this hardware redundancy to support a system capable of operating even if the driver is asleep, without requiring sudden handovers.
Why It Matters
Anderson’s remarks highlight a growing philosophical divide in the autonomous vehicle sector between camera-centric developers like Tesla and those advocating for sensor fusion. Tesla has faced criticism for delays in achieving full autonomy with its vision-only approach, while other developers, including Waymo, have publicly stated that cameras alone are insufficient. GM’s strategy for its 2028 launch explicitly rejects the "Level 3" model of dynamic handovers, aiming instead for a "graceful countdown" that allows drivers 10 minutes to retake control before the vehicle safely pulls off the highway.
The Bottom Line
GM is doubling down on a multisensor approach for its 2028 autonomous highway feature, with executive leadership framing the removal of radar as a technically flawed decision that prioritizes ideology over safety and physics.
