Waymo Unveils 6th-Gen Driver: Scaling Autonomous Mobility with Lower Costs

Waymo's 6th-generation Driver is a cost-effective autonomous system designed for rapid expansion into new cities and diverse weather conditions. It features a sophisticated sensing suite and custom silicon that delivers higher performance with fewer hardware components. This platform-agnostic technology is now entering fully autonomous operations, paving the way for high-volume production and broader public access.

Key Points

• The 6th-generation Waymo Driver reduces hardware costs through a streamlined configuration and custom silicon while improving performance and safety.
• A multi-modal sensing suite combining high-resolution 17MP cameras, lidar, and imaging radar allows the vehicle to navigate complex environments and extreme weather.
• The system is designed for versatility across different vehicle platforms and is engineered for high-volume manufacturing of tens of thousands of units per year.
• Advanced External Audio Receivers (EARs) enable the system to detect and localize emergency sirens and railroad crossings before they are even visible.
• Waymo is transitioning to fully autonomous operations with this new generation on the Ojai platform, starting with employee and guest testing.

Sentiment

The community is predominantly supportive of Waymo's technical approach and skeptical of Tesla's vision-only strategy. Most commenters view multi-modal sensor fusion as the right path to safe autonomous driving. Tesla's robotaxi program faces intense skepticism due to Elon Musk's long track record of broken self-driving promises. However, there is a vocal minority defending Tesla's cost and scalability advantages, making the overall tone one of qualified enthusiasm for Waymo tempered by real questions about cost and geographic scalability.

In Agreement

• Multi-modal sensor fusion with lidar, radar, and cameras provides essential redundancy that vision-only systems cannot match, especially in adverse weather, nighttime, and edge cases
• Waymo's nearly 200 million autonomous miles of experience gives it a nearly insurmountable lead in perception technology, comparable to NVIDIA's CUDA advantage
• The platform-agnostic 6th-generation design is a smart scalability strategy, allowing Waymo to deploy across multiple vehicle platforms
• Lidar costs are dropping rapidly, with Chinese manufacturers approaching sub-$200 units, undermining the economic argument against multi-sensor systems
• Waymo's perception technology could generalize beyond ride-hailing to robotics, factory automation, and alternative vehicle forms for last-mile transportation
• Tesla's decision to abandon lidar and radar is increasingly seen as a strategic mistake that will be hard to reverse

Opposed

• Vision-only systems are cheaper and more scalable, allowing data collection from millions of consumer vehicles rather than a small fleet, which could ultimately win on cost per mile
• Waymo is not truly 'fully autonomous' since it relies on remote fleet response operators to provide navigational guidance in complex situations
• Waymo's geographic limitations and fleet size mean it cannot compete with companies that can deploy autonomous driving globally on consumer vehicles
• Sensor fusion adds complexity and potential failure modes—some engineers argue it's often better to master one sensor type than try to reconcile conflicting data from dissimilar sources
• The US should invest in mass transit rather than doubling down on autonomous cars, which perpetuate car-dependent infrastructure
• Waymo's approach may win for now but could be overtaken by cheaper vision-based systems as AI and computer vision continue to improve rapidly