Unsupervised Tesla robotaxi fleet expands as company reports rising vehicle count

Tesla’s robotaxi story is now split in two. On one side, the fleet is clearly getting bigger — more driverless Model Ys are showing up in Texas cities, and Tesla has already told investors Dallas and Houston went live in April. On the other side, the service still looks thin, slow, and tightly constrained in actual use. That gap is the whole story right now. The news on May 13 is that the expansion is real, but so are the limits. ### What actually expanded? The clearest update came from fleet tracking compiled by Tesla watchers and picked up on May 13: unsupervised vehicles increased in Dallas and Houston, pushing the visible Texas total to about 38. The rough city split being cited is 27 in Austin, 5 in Dallas, and 6 in Houston. Tesla itself has not published a live public fleet dashboard, but it did confirm in its Q1 2026 update that unsupervised robotaxi rides launched in Dallas and Houston in April, adding to the Austin operation. (cleantechnica.com) ### Why does “unsupervised” matter? Because this is the part investors care about most. A supervised test car still has a human ready to take over. An unsupervised robotaxi is the commercial claim — the car is doing the driving job on public roads without a driver at the wheel. That does not mean full open-city autonomy. It means Tesla is confident enough to run a limited service in specific geofenced areas under specific rules. (cleantechnica.com) Basically, the milestone is real, but it is narrower than the branding makes it sound. ### So why are riders still complaining? Because a bigger fleet does not automatically mean a good service. Reuters’ testing in Dallas and Houston found long waits, poor availability, and drop-off points that could still leave riders far from where they wanted to go. One Dallas trip that would normally take around 20 minutes by car turned into nearly two hours from request to arrival because of high demand and delays. That is not just annoying — it tells you the network is still sparse and heavily constrained. (assets-ir.tesla.com) ### Why are the routes so weird? Turns out a lot of the “convenience” problems look like safety limits in disguise. Electrek’s read on the Reuters test is that Tesla is avoiding harder driving situations by sticking to surface streets, limiting operating zones, and keeping availability low. If the system struggles with certain turns, speeds, merges, or traffic patterns, the easiest near-term fix is not better service — it is shrinking the problem. (money.usnews.com) That keeps the cars running, but it also makes the product feel unfinished. ### Is regulation about to get tighter? Yes — at least in Texas commercial operations. The Texas DMV says that starting May 28, 2026, companies operating automated vehicles commercially on Texas roads need an active authorization. The rule applies to Level 4 or Level 5 automated driving used to transport passengers or property without a human driver. So Tesla is expanding right before the state moves from a lighter-touch posture to a formal authorization regime. (electrek.co) ### Does this mean Tesla is ahead or stuck? A bit of both. Tesla is ahead in the sense that it has a real unsupervised service running in multiple Texas cities, which is more than a slide deck or a demo loop. But it still looks stuck on the hard part — scaling from “cars exist” to “riders can depend on this.” Fleet count is the easy headline. Reliability is the real test. (txdmv.gov) ### What should people watch next? Watch three things. First, whether the visible fleet count keeps rising after May 28. Second, whether wait times and routing improve instead of just vehicle totals. Third, whether Tesla broadens geofences or keeps the service boxed into easy territory. If those service metrics do not improve, more cars alone will not prove readiness. ### Bottom line? (assets-ir.tesla.com) Tesla’s robotaxi network is expanding — that part looks real. But the current version still behaves like a carefully managed pilot, not a city-scale transportation service. More unsupervised cars matter. The catch is that the service problems matter more. (cleantechnica.com) (txdmv.gov)