There is a very special partner of EVwire, without whom we couldn’t have started this reporting at all.

We’ve been working together with Voltera for years now, and it is one of the core reasons why I went down that robotaxi rabbit hole in the first place — they are working with the key players in the space. They fall into the layer of industry that I call the robotaxi “enablers”.

Here’s a quick intro in Voltera’s own words:

“Voltera develops and operates charging infrastructure engineered for autonomous mobility.

Our sites deliver reliable high-power charging in key geofenced zones, helping robotaxi fleets maximize uptime and scale from pilot programs to full commercial deployment.”

Important: this newsletter does not go through any editorial overview of Voltera’s team, nor can they influence any of the reporting — the words in this newsletter are mine alone.

Let’s take a step back.

I wasn’t lying, nor hyping, when I said there were a million paid robotaxi rides done this week. There were over a hundred thousand robotaxi rides provided on the planet today.

🇺🇸 About half of the one million robotaxi rides this week were done in the US.
The largest player, Waymo, is now executing at 450k+ paid rides per week.

🇨🇳 Most of the other half of that million robotaxi rides took place in China.
Baidu, the largest on the other side, is doing 250k+ paid rides per week.

🇦🇪 The Middle East is becoming a hub for deployments. Already ahead of Europe, with multiple fully autonomous fleets roaming the streets of Abu Dhabi, Dubai, etc. The players featured here will be most of what we see in China.

🇪🇺 …Europe? Europe is just starting up. But quite massively in 2026.
It’s a convergence point in 2026 for the US and China side robotaxi players, with a few (well, two) domestic players slightly earlier in the game.

The fleet sizes in Europe will be small for this year as the services are still starting (Waymo, for example, has a 30+ fleet in London roads right now, testing).

I left the partnerships with ride-hailing companies (Uber, Lyft, Bolt) and other connections off these images on purpose for now. But we will talk a lot about them in the next newsletters.

While exact numbers aren’t often published as things move fast, here’s what we know as of today on the bigger players:

Now, these are the ‘top five’ in terms of actual robotaxis on the roads today.

For quite a few players, from strong China makers to Tesla, we’re right on the cusp of them scaling up the fleet. But current operations are showing the reality, since rolling out the real robotaxi service is, quite obviously, not just about the cars.

We’ll tackle the two regions most of our 14,000 readers are from: the US & Europe. In China, you can hail several robotaxi players in at least the 5 major cities.

We’ve got several strong players here.
Rollout is currently fully dominated by Waymo in the US. So we’ll start there.

You can catch a fully driverless Waymo through their app (or through Uber in 2 cities) today in:

And in all these cities below, Waymo is already running its fleet to get ready, either with a safety driver in the early stages of development, or already driverlessly for employee-only testing ahead of commercial roll-out:

I also enjoy the way Waymo announces each new city with a little personal touch:

Waymo plans to be in “20 or more” cities by the end of this year, and 10 are live today. Some are deployed with partners (e.g., with Lyft in Nashville, Tennessee), but most currently run solo on Waymo’s ordering side.

The partners (whom I call “enablers”) on the back side, from the fleet charging depots to maintenance, vary per market.

Now, let’s look at one that runs fully solo, from making the vehicles, operating the service, to charging and maintaining them: Tesla.

We have already seen Tesla launch the commercial operations in the San Francisco Bay Area and in Austin, Texas.

Now, the core robotaxi network for Tesla has always been meant to be the two-seater Cybercab (we’ll cover why the two-seat configuration for the base fleet makes a lot of sense sometime later), and the mass production of the Cybercab has been set to start from April, which is when we’ll expect the rollout ramp-up.

There have already been 25 Cybercabs spotted testing in Austin, 3 in SF, and more across the US. And the Giga Texas flyovers show the company seems to be well on track to start mass production soon.

Here are the cities that Tesla has confirmed, on their Q4 earnings call, that will get the public Robotaxi service in the first half of 2026:

As you can see, all these cities are also where Waymo already operates or plans to. This is no accident — even Elon has hinted that Waymo is paving the legislative way for them:

The benefit for Tesla’s approach is that once it has the tech side (actual Full Self Driving as we’ve seen with its few true robotaxis in operations now) and the legislative side in any market ready, it can flood those markets with ready-from-the-factory cybercabs at an almost ridiculous volume.

Zoox, the purpose-built robotaxi company owned by Amazon, is now live in limited areas of Las Vegas and San Francisco, and launching soon in Austin and Miami.

Due to the ownership of the company, I’d bet the company must be working on a fully autonomous delivery form factor of the Zoox vehicles as well (the shape seems fitting already), although there’s no official communication about it. Yet?

I’m leaving the ‘coming soon’ of other players, like Lucid’s Nuro x Uber deal; Motional planning to launch driverless in Las Vegas, and the VW ID Buzz deal out of this list for now — but there’s a lot of news coming soon.

In the US, Uber is currently leading in terms of partnerships with nearly all robotaxi players out there, hoping to become the key facilitator of robotaxi rides worldwide:

For example, just this week, Uber announced another one for this blue box there: a multi-year partnership with Zoox, starting this summer in Las Vegas and next year in Los Angeles, Uber riders can get matched with a Zoox robotaxi.

Lyft has its own set of partnerships & operations lined up for robotaxis, too, which we’ll dive into when the time is right.

Nearly nowhere. There are some shuttle-type services live today, but those are currently operated mostly as pilot services, and there’s no built-out robotaxi service quite yet.

But here’s what’s coming from the larger players this year:

In addition to these, we’ve also seen Bolt, the ride-hailing giant this side of the pond, partnered with Pony AI to launch robotaxis on European streets, and also with Stellantis, creating one of the rare Europe-Europe robotaxi partnerships (even though Stellantis itself relies on outside tech like NVIDIA for the self-driving side). Bolt targets 100,000 autonomous vehicles on the platform by 2035.

In fact, surprisingly the ride-hailing platforms like Bolt, Uber, Lyft et al have become very interesting to watch, as they work through on how to become the default demand layer for the robotaxis, without becoming AV labs themselves.

Trials start this year, industrial scale-up target around 2029. Meanwhile, Stellantis and Pony.ai are real-world testing the autonomous Peugeot e Traveller vans in Luxembourg and plan to deploy them in several cities from 2026 onward.

It sure won’t be boring around here.
And this newsletter right here is how you’ll be able to track the progress.

US policy swing towards robotaxis is in full … swing. The first Autonomous Vehicle Safety Forum was held this week at USDOT Headquarters in DC. Here’s a direct quote from Sean Duffy:

RESOURCE TIP: You can find the full recordings of all event panels and remarks on the NHTSA site here.

In every newsletter, we’ll look into what is happening with robotaxis on the streets, both the good and bad. Keep in mind, that we will see also edge cases and some obvious misbehaviors of the robotaxis that I will also always showcase.

Different players in the industry are at different stages of technological progress, using different sensor suites and models for training their service. And of course, the ‘normal’ rides never make the news. Since we should strive to a maximally safe system, including being safer than most human drivers, I think it is equally important to call out the shortcomings.

Today, we’ve got room for two specific cases on railroads:

Video: Waymo performs an unprotected left turn across several lanes, and… stops in the middle. A very, very dangerous situation. The other problematic incident we just saw on March 7th was with this Waymo stopping on the path of of the railroad crossing barriers, which ended up resting on its roof. The train narrowly missed the vehicle.

In a statement, Waymo said there were no passengers in the car, and the vehicle behaved safely in this scenario, stopping about eight feet from the tracks.

"Waymo vehicles have safely traversed railroad crossings millions of times fully autonomously. In these instances, the vehicles performed a controlled stop more than 2.5 meters before the tracks, after the crossing lights activated as the vehicles approached the crossing. […] While the vehicles remained at a distance from the tracks, we recognize the concern caused by stopping past the crossing gate.”
— Waymo spokesperson statement to The Independent

Waymo said it was temporarily restricting its vehicles’ travel at similar crossings while it conducted a thorough analysis of the incident and developed the stopping system’s response.

Video: The railroad crossings seem to be an edge case yet to be fully solved, as a user video from March 9th shows Tesla’s FSD (Supervised) — which, granted, isn’t a robotaxi just yet and is meant to be a driver-supervised system — drove straight through the same kind of railroad crossing arms.

There’s no info currently as to which version the Tesla was running on. This can be surprisingly important in determining the current capabilities and safety, and you’ll learn why in our upcoming newsletters.

Got a good video, interesting situation (good or bad) that we should cover in our newsletter next week? Drop me a note.

We’ll start with two a great resources in the robotaxi space that you can check out right away, which is offering visual tracking on the robotaxi rollouts:

AVmap.io

It is a resource originally built by Jackson Lester, and the ever-expanding dataset covers 15 companies across roughly 60 cities in 13 countries, with historical deployment data back to April 2017. The data is open source, available through github, looking for contributors as well.

You can basically watch the robotaxi service areas grow as if looking at a petri dish.

Now, this resource gained enough traction to catch the eye of Waymo, which went and hired Jackson as Product Manager two months ago.

Since you can’t really run a public competition tracker when working at one of these yourself, the AVmap found a new home at Institute of Transportation Studies Berkeley (ITS), which will keep updating it.

Next newsletter, we’ll cover another great tracker that ended up propelling its founder to employment in one of the robotaxi companies.

I guess if you’re looking for one the robotaxi industry jobs… you know what to do ;)