From 14 cities to 15,000: What will it take for robotaxis to scale?

Robotaxis are proven. Waymo has logged more than 220 million autonomous miles and, by its calculations, has had more than 94% fewer serious injury crashes than human drivers. It provides about half a million paid rides per week and is expanding to more cities.

So the question has shifted from “Does it work?” to “How does it get anywhere?”

Waymo is present in 14 cities. Uber is among more than 15,000. Uber took 3.9 billion trips last quarter. It processes Waymo’s entire volume of weekly trips every 17 minutes. Driving software and vehicle supply have billions stacked against them. Almost no one funds the least fascinating problem: what happens to cars between races.

The pit is on the other side of town

No Formula 1 team goes back to the factory for new tyres. The pit is on the track; the car comes out in a few seconds. A robotaxi finishes a trip, discovers spilled milk in the back seat, and drives up to 15 miles, empty, to a centralized depot to be cleaned, loaded and inspected.

Each restore costs approximately two hours. California data shows the cost: Of 86 million Waymo miles driven through 2025, only 54% carried a passenger.

Not every empty mile is a depot run, but every depot run is an empty mile. Autonomous on the road, manual when it requires maintenance. This is where the scale breaks.

It worked for a company in some cities. It cannot bring an industry to thousands of people, for three reasons.

The earth is no longer there. The site that needs a warehouse, industrial area and electrical wiring, is what everyone wants. Brokers report that data center developers have outbid industrial buyers for powered packages. An operator entering a new city participates in a home auction against the likes of Microsoft and Amazon.

The clock is against you. The California regulator is targeting six months just to power a standard charging site. A new circuit lasts 1.9 years; a substation upgrade, 2.8; a new substation, 8.9. By city. In thousands of cities, that is decades.

Every city is different. Phoenix is ​​a sunlit expanse of parking lots. Zurich is a medieval street plan with a planning committee. Riad has heat that cooks the batteries and dust that covers the sensors at lunchtime.

There is no single project that fits all, so the industry builds a bespoke depot for everyone. Fifteen cities, 15 projects. Fifteen thousand, and the model collapses under its own paperwork.

Waymo is the beginning, not the form

Waymo did it all themselves. He designed the driver, adapted the cars, managed the fleet, created the app, staffed the depots. Waymo also dropped two tiers: Zeekr builds its Ojai vans, while Moove, Avomo, Avis and Flexdrive operate its depots.

The next wave of autonomous vehicles will be built by specialists. Wayve, Nuro and Mobileye sell brains. Automakers put these brains in their cars: Nissan with Wayve, Lucid with Nuro, Volkswagen with Mobileye, Mercedes with NVIDIA. Uber supplies passengers to more than 30 independent partners. The result is not a fleet of just one vehicle, but dozens of fleets of different vehicles.

I spent years building a battery swap network for shared scooters, when hundreds of operators operated vehicles from a few manufacturers in thousands of cities. Nobody won because they had a better scooter. Survivors could load, repair, and relocate tens of thousands of vehicles per day. The vehicle was never the difficult part. The operations were.

Robotaxis are the same thing at a hundred times the price. A Wayve-powered Nissan on Uber in Zurich still needs to be charged, cleaned and inspected several times a day, and no one wants to buy an industrial lot there. Without the infrastructure, nothing works.

What the unlock scale for robotaxis looks like

The container did not make ships faster. It made all ports equal and this unlocked global trade. Robotaxis need their own container, a standard unit of service infrastructure that fits in a bay or two, sets up in days, not years, is configured by city, not redesigned, and serves any fleet on any network, as one cell tower serves every carrier.

Full disclosure: My company, Aseon Labs, is building a robotic charging, cleaning and inspection station that arrives on a flatbed, designed to go live within a day. We are early and we won’t be the only ones. The topic is not up to us. It depends on arithmetic: years per depot, per city, for an industry that needs thousands of them.

Every robotics demo ends the same way: someone off-camera walks in with a charging cable. Robotaxis are the first commercial-scale robots to learn that the demo was never the hard part. Andreessen Horowitz raised $1.1 billion for physical development of artificial intelligence. Software has eaten the world. Now he needs a place to park.

Cars can already drive themselves. They can’t load, clean or inspect themselves yet. Someone has to build the pit.

About the author

Dan Keene is co-founder of Aseon Labs, a Y Combinator-backed company that builds distributed robotic gas stations for fleets of autonomous vehicles.

He previously co-founded Pushme, the battery swap network acquired by TIER in 2020.