Waymo Deploys 6th-Gen Hardware to Public Roads: No Safety Drivers, Tens of Thousands of Units Per Year
The robotaxi economics inflection point may have finally arrived
By Russ Calder | February 12, 2026
Waymo announced today that its 6th-generation autonomous driving system is now operating on public roads without safety drivers. The new hardware, manufactured at scale in Phoenix, represents the company's most aggressive push toward commercial viability: fewer sensors, higher fidelity, and a production target of "tens of thousands of units per year."
This isn't a prototype unveiling or a pilot expansion. Waymo is deploying cost-optimized hardware at industrial scale while maintaining fully autonomous operations. After seven years of safety-validated service and 200 million miles driven without human backup, the company is betting that its economics have finally caught up to its technology.
The Cost Breakthrough
The autonomous vehicle industry has spent a decade wrestling with one fundamental problem: the sensor stack costs too much. Early self-driving cars carried $75,000 to $150,000 in lidar, radar, cameras, and compute hardware. At those numbers, the math for profitable robotaxis simply doesn't work.
Waymo's 6th-gen system attacks this problem directly. The new configuration uses less than half the number of cameras compared to the previous generation while dramatically increasing sensor quality. The headline specification: 17-megapixel imagers, compared to the 5-8 megapixel cameras standard across the industry.
The strategy is counterintuitive but sound. Instead of layering more sensors for redundancy, Waymo invested in fewer, better sensors paired with custom silicon. The result is a system that costs less to manufacture while capturing more visual information per frame.
"Demonstrably safe AI requires equally resilient inputs," Waymo stated in its announcement. The company's position: higher-fidelity sensors allow AI models to make better decisions, reducing the need for overlapping coverage from multiple lower-quality cameras.
Manufacturing at Scale
The Phoenix factory represents the real story. Waymo has historically operated hundreds of vehicles across its service territories. "Tens of thousands of units per year" suggests a 10x to 100x expansion in manufacturing capacity.
This production target matters for two reasons. First, it signals that per-unit costs have dropped enough to justify volume production. Second, it indicates Waymo believes it can deploy vehicles faster than regulatory and market expansion will limit growth.
The company is taking a platform-agnostic approach, describing itself as "building a Driver, not a vehicle." The 6th-gen system will run on both Waymo's custom Ojai platform and the Hyundai IONIQ 5. This flexibility allows Waymo to scale across different vehicle partnerships without redesigning its sensor integration for each platform.
Initial deployment is cautious. Waymo is rolling out 6th-gen hardware on Ojai vehicles first, running employee and guest trips while refining the rider experience. Public availability will follow, though the company hasn't committed to a specific timeline.
Technical Specifications
The 6th-gen sensor suite includes cameras, lidar, imaging radar, and external audio receivers. Each component received significant upgrades:
Cameras: The 17MP imagers provide roughly 2-3x the resolution of industry-standard sensors. Higher resolution means the system can identify objects at greater distances and with more certainty, particularly important for detecting pedestrians, cyclists, and road debris.
Lidar: Waymo uses industry-standard lidar hardware with custom optics and processing silicon. This approach reduces manufacturing costs while maintaining the precise depth mapping that lidar provides.
Radar: The new imaging radar system performs better in adverse conditions, specifically heavy rain, snow, and roadspray. This capability is essential for geographic expansion beyond the dry climates of Phoenix and most of California.
Audio: External microphones help the system detect emergency vehicles and other acoustic signals that cameras and radar might miss.
The multi-modal approach remains Waymo's fundamental bet: that combining different sensor types produces safer outcomes than any single modality alone.
Winter Weather and Geographic Expansion
The improved radar and sensor suite opens new markets. Waymo has operated primarily in Sun Belt cities where weather conditions are predictable. San Francisco's fog tested the system, but true winter weather (snow, ice, heavy rain) remained a limitation.
The 6th-gen hardware explicitly targets these conditions. Heavy rain obscures cameras and scatters lidar returns. Snow covers lane markings and changes road surfaces. Roadspray from other vehicles creates visibility challenges that compound in traffic.
Waymo claims the new system handles these scenarios. If validated, this opens deployment possibilities in Chicago, Boston, Seattle, and other major metropolitan areas where winter conditions previously posed obstacles.
The company currently operates in San Francisco, Phoenix, Los Angeles, Austin, and several other cities. Geographic expansion announcements typically follow hardware deployment by 6-12 months as Waymo maps new service areas and navigates local regulatory requirements.
Competitive Context
Waymo's announcement lands in a competitive environment that has shifted significantly over the past year.
Tesla continues to promise unsupervised Full Self-Driving capability but relies on a vision-only approach. Elon Musk has argued that cameras are sufficient, pointing to human drivers who navigate using only visual input. Tesla's advantage is scale: millions of vehicles collecting training data versus Waymo's thousands.
Cruise, formerly Waymo's closest competitor in commercial robotaxi operations, paused service in late 2023 following an incident where a vehicle dragged a pedestrian. The company has not resumed fully autonomous public operations.
Zoox, Amazon's autonomous vehicle subsidiary, continues testing but hasn't launched commercial service.
This leaves Waymo with a meaningful first-mover advantage in deployed, revenue-generating robotaxis. The 6th-gen announcement extends that lead with hardware designed for cost-effective scaling.
What to Watch
Three metrics will determine whether this announcement represents a genuine inflection point:
Production numbers: Waymo said "tens of thousands" but hasn't provided specifics. Actual manufacturing output over the next 12 months will reveal whether the Phoenix factory can deliver.
Geographic expansion: New city announcements, particularly in winter-weather markets, will test the improved sensor claims.
Unit economics: Waymo doesn't disclose per-ride profitability. Industry observers will look for signals in pricing, service hours, and fleet utilization rates.
The autonomous vehicle industry has absorbed billions in investment and produced modest commercial results. Waymo's 6th-gen hardware represents a clear thesis: the technology is mature, the costs are manageable, and the scale is achievable.
The next 12 months will test whether that thesis holds.
Russ Calder covers robotics and embodied AI for Synthetic.
