Tesla’s making big moves with its robotaxi service, pushing past old limits as it races to catch up with rival Waymo. The company announced a 50% increase in its Austin robotaxi fleet size in August 2025, though it hasn’t revealed the original fleet numbers. Even with this expansion, Tesla’s fleet remains about one-third the size of Waymo’s operation.
Tesla’s robotaxi fleet grows 50% in Austin, still trails Waymo’s larger autonomous operation
The robotaxis run on Tesla’s Full Self-Driving software and use eight cameras to capture a complete view around the vehicle. Unlike Waymo, Tesla doesn’t use radar or lidar sensors. Instead, the company relies entirely on cameras and artificial intelligence to process visual data. This approach has drawn criticism from experts who say camera-only systems aren’t reliable enough for long-term operations. Elon Musk claims FSD version 14 will be 2-3 times safer than human drivers, with future versions potentially reaching even higher safety multiples.
Tesla’s system still requires human supervisors in the vehicles during operations. These safety monitors can intervene using an In Lane Stop button on the touchscreen when needed to prevent collisions. This contrasts sharply with Waymo, which operates fully autonomous vehicles without any human drivers. Waymo currently provides about 250,000 paid rides per week with completely driverless cars. Industry experts estimate Tesla’s about five years behind Waymo’s autonomous capabilities.
Despite the gap, Tesla’s FSD system shows impressive abilities in some areas. The technology maneuvers through complex traffic situations without hesitation and makes left turns decisively. The system’s supercomputer enables consistent performance across different countries. It can select ideal paths during challenging routing scenarios, sometimes performing better than human drivers in certain situations. Tesla’s autonomous systems benefit from continuous data collection spanning over nine billion miles of real-world driving experience.
However, problems persist. The robotaxis experience phantom braking incidents where vehicles suddenly stop without visible obstacles. These events have caused passenger belongings to fly around inside the cars. A Carnegie Mellon University professor identified phantom braking as a design flaw in the system. These issues stem from AI hallucinations, which experts say are a natural part of machine learning operations.
Tesla’s following regulatory pathways already established by competitors in US markets. The company’s considering Manitoba, Canada as a potential testing location for future launches. International implementation plans depend on getting regulatory approvals in target areas.
The real competition with Waymo won’t begin until Tesla achieves unsupervised autonomous operations. Until then, Tesla’s robotaxis remain a supervised system that’s rapidly expanding but still catching up to established competitors.