Tesla has expanded the geofenced area of its invite-only robotaxi service in Austin, Texas, surpassing rival Waymo’s footprint in the city.
The company began offering paid, fully driverless rides in Austin in late June with a safety operator always seated in the passenger seat.
As a start, the company began with a fleet of about 10 Model Y vehicles covering a five-mile radius.
Since then, Tesla has expanded the coverage area three times and introduced dynamic pricing based on distance, replacing a fixed fare structure.
Tuesday’s expansion brings both Tesla’s Giga Texas factory and Austin-Bergstrom International Airport inside the service zone. The new geofenced area has about 170 suare miles.
The rides continue to operate with a supervisor seated in the passenger seat, unlike Waymo’s cars, which run without human monitors and are open to the general public.
Tesla’s expansion comes weeks after the U.S. Department of Transportation and the National Highway Traffic Safety Administration (NHTSA) confirmed it had contacted the company in July, requesting a ride and details on its Austin pilot.
The agency sought information on the service’s operational boundaries, maximum operating speeds, future plans for public use, and whether in-vehicle supervisors or remote operators would remain in place.
NHTSA said Tesla submitted responses, which were treated as confidential.
Tesla also launched a robotaxi pilot in the San Francisco Bay Area late last month. Due to local regulations, those rides still use a safety monitor in the driver’s seat.
Meanwhile, Waymo integrated its service into Uber’s app in Austin earlier this year and has begun pilot rides in Atlanta, with launches planned in Miami and Washington, D.C. next year.
As reported by EV, Tesla chief executive Elon Musk reiterated on Monday his opposition to LiDAR sensors, following comments by Uber CEO Dara Khosrowshahi, who called the technology “crucial for safety” and said autonomous vehicles must be “multiple times better than humans.”
Musk countered that Tesla’s camera-only approach avoids conflicts that he argues arise when multiple sensors are combined.
