Waymo says cameras aren’t enough for full autonomy; Waymo vs. Tesla sensor debate

What happened

• Waymo’s VP of Onboard Software, Srikanth Thirumalai, stated that cameras are not enough to achieve full autonomy.
• After more than 200 million real-world miles, Waymo concluded that safe, fully autonomous operations require more than just cameras.
• Waymo uses cameras along with lidar and radar to create a ‘rich, redundant world view’ that no single sensor can replicate.
• Tesla’s head of AI, Ashok Elluswamy, argued that multi-sensor approaches like lidar and radar are not necessary for safe autonomy; cameras alone can provide safe autonomy.
• Waymo operates its robotaxis in 11 cities, conducting 500,000 paid trips a week.

Why it matters for me

This discussion highlights the fundamental difference between relying solely on cameras and adopting a multi-sensor approach (like lidar and radar) for autonomous vehicles. Waymo believes that true Level 4 autonomy requires a purpose-built system validated by real-world experience, not just improving current driver-assist systems.

What to remember

• True L4 maturity in autonomous driving requires a purpose-built system, validated on closed courses and hardened by real driving experience.
• The debate centers on whether cameras alone are sufficient or if sensor redundancy is essential for safety at scale.
• Tesla’s current FSD system is classified as Level 2, meaning drivers must constantly monitor the vehicle.

Source: Original article