Tesla just skipped a step it has taken in every other city. On July 3, the company launched Robotaxi service in Miami with no safety monitor in the car from the very first ride — no human in the driver's seat, no one in the back watching over the system. It's the first time Tesla has done this anywhere outside Texas, and the first time it's done it in a brand-new market with zero transition period.
That's a meaningful jump in how much trust Tesla is putting in its own software, and it's worth understanding exactly what changed and what didn't.
What actually launched
The service covers a geofenced zone of roughly 10 to 14 square miles across West Miami, Doral, and Coral Gables — deliberately excluding downtown, Miami Beach, and the airport terminals. Riders book through Tesla's dedicated Robotaxi app on iOS or Android, though the company has told users to expect a waitlist, the same rollout pattern it used in Austin, Dallas, and Houston.
Tesla's VP of AI Software, Ashok Elluswamy, confirmed on social media that the cars are running with no safety monitor and no driver at all — a detail quickly backed up by rider videos showing empty front seats. The entire fleet currently runs on Model Y vehicles. Tesla's purpose-built, pedal-free Cybercab has only just started limited public-road testing in Austin and isn't expected to join active markets until production scales later this year.
Here's what makes Miami different from Tesla's earlier launches: when the company brought Robotaxi to Austin in June 2025, every car carried a human safety monitor in the passenger seat for months before Tesla pulled them out. Dallas and Houston, which launched in April, followed a similar phased approach. Miami skipped that transition completely — driverless from day one, in a state and a city Tesla has never operated in before.
Why Florida, and why now
Florida has one of the most permissive regulatory environments in the US for autonomous vehicles — it requires no AV-specific state permit, which made it a logical next stop. Musk has framed the decision to skip the safety-monitor phase as a sign of growing confidence in Full Self-Driving software following more than a year of real-world data collection across Tesla's existing markets.
But Miami is also a genuinely different test than anything Tesla's system has handled at commercial scale. Summer in South Florida means sudden, heavy downpours, sharp sun glare, and high humidity — conditions Tesla's camera-only approach hasn't previously had to manage in live commercial operation. That matters because Tesla's Full Self-Driving system relies entirely on cameras, with no lidar or radar backup. Rivals like Waymo use a mix of lidar, radar, and cameras, arguing that sensor redundancy provides a meaningful safety margin exactly in conditions like heavy rain or glare, where camera-only systems are more likely to struggle.
This isn't a hypothetical concern. In March, the National Highway Traffic Safety Administration escalated a probe into Tesla's Full Self-Driving system to an engineering analysis — the step that typically precedes a possible recall — after finding the camera-only system fails to detect or warn drivers appropriately under degraded visibility conditions like glare and airborne obscurants.
The scale gap is still enormous
Here's the context that often gets lost in robotaxi headlines: Tesla being first to a city doesn't mean Tesla is winning on scale. Texas DMV filings released in late May showed Tesla had just 42 autonomous vehicles authorized for driverless ride-hailing in the state — less than one-tenth the size of Waymo's Texas fleet of 577 vehicles. Nationally, Waymo operates an estimated 3,000 robotaxis across more than 10 US metro areas and is reportedly targeting a million paid rides per week.
| Metric | Tesla Robotaxi | Waymo |
|---|---|---|
| US cities live | 4 (Austin, Dallas, Houston, Miami) | 10+ metro areas |
| Texas authorized fleet | 42 vehicles | 577 vehicles |
| Approx. national fleet | Low hundreds | ~3,000 vehicles |
| Sensor approach | Camera-only (vision) | Lidar + radar + camera |
| Miami safety monitor? | None, from day one | None (established market) |
| Published safety data | Not disclosed for unsupervised ops | Regularly published |
Sources: Texas DMV filings, CNBC, TechTimes, Electrek
Waymo has also already announced plans to launch in Miami itself, along with Atlanta and Washington, D.C., meaning the two companies' driverless playbooks are about to compete head-to-head on the same Florida streets. Waymo has a considerable head start there too — it began driverless service in Miami back in April, months before Tesla's launch.
Tesla has not published safety data specific to its unsupervised operations in any market it currently serves. The only mandatory disclosure tied to these vehicles is federal crash reporting under NHTSA's standing general order, which documents incidents after they happen rather than offering safety evidence ahead of deployment. Tesla's own NHTSA filings through April 2026 document 17 incidents across Austin operations — 13 property damage only, two with no injuries, and two involving injuries, one requiring hospitalization. Notably, all 17 of those incidents involved supervised vehicles with a safety monitor present; none involved the fully unsupervised fleet, though the unsupervised fleet is also far smaller and has logged far fewer total miles.
What this means if you're not in Miami
Even if you'll never ride in one of these cars, the Miami launch matters as a signal. Tesla has now removed human oversight in four cities — Miami, Dallas, and Houston fully, with Austin running a mix of supervised and unsupervised vehicles. The Bay Area remains limited to monitored rides only. If Miami's unsupervised service holds up through a full Florida summer without a spike in serious incidents, it answers one of the central technical questions skeptics have been raising about camera-only autonomy for years.
If it doesn't, the fallout — regulatory, reputational, or both — will likely shape how fast every other AV company is allowed to move next.
My Take
[Personalize this section before publishing — starting draft below.]
What stands out to me isn't that Tesla launched in a new city — it's that it launched with zero transition period, in a market with weather conditions its system has never handled at commercial scale, while a federal safety probe into that exact failure mode is still open. That's either the confident move of a company that has genuinely solved the hard part, or a bet that outruns the safety data backing it up. Right now, there's no published evidence to tell us which. I'd want to see a full rainy season of real Miami operation, with actual incident numbers, before calling this a win either way.
FAQ
Does Tesla's Robotaxi in Miami have a safety driver?
No. Tesla confirmed the Miami service is running fully unsupervised from launch, with no human in the driver's seat and no safety monitor anywhere in the vehicle.
How does Tesla's Robotaxi fleet compare to Waymo's?
Tesla trails significantly. Texas DMV filings show Tesla has 42 authorized vehicles in the state versus Waymo's 577. Nationally, Waymo operates an estimated 3,000 vehicles across 10-plus cities, while Tesla's fleet remains in the low hundreds across four cities.
What areas of Miami does the Tesla Robotaxi service cover?
The initial geofenced zone spans roughly 10 to 14 square miles across West Miami, Doral, and Coral Gables. Downtown Miami, Miami Beach, and the airport are not yet included.
Why is Tesla's camera-only approach controversial?
Tesla's Full Self-Driving system relies entirely on cameras, with no lidar or radar. In March 2026, NHTSA escalated a probe into whether this approach fails to properly detect hazards in low-visibility conditions like glare and heavy rain — conditions Miami's summer weather produces regularly.
Has Tesla published safety data for its driverless rides?
No. Tesla has not released safety data specific to its unsupervised operations in any market. The only public disclosure is federal crash reporting through NHTSA, which reports incidents after they occur.
