Reset downtime: Each robotaxi reset costs roughly two hours, with depot runs of up to 15 miles driven empty.
- Proven autonomy, unproven logistics: Waymo has logged 220 million autonomous miles with 94% fewer serious injury crashes, but only 54% of its 86 million reported California miles carried a passenger.
- Grid bottleneck: Charging infrastructure timelines run from 1.9 years for a new circuit to 8.9 years for a new substation.
- Fragmenting market: Zeekr, Moove, Avomo, Avis, Flexdrive, Wayve, Nuro, Mobileye, Nissan, Lucid, Volkswagen, Mercedes, and NVIDIA are splitting the robotaxi stack into specialized layers.
- Distributed servicing bet: Aseon Labs, backed by Y Combinator, is building a flatbed-delivered robotic charging, cleaning, and inspection station designed to go live within a day, as Andreessen Horowitz commits $1.1 billion to AI's physical buildout.
The robotaxi question has shifted. It is no longer whether autonomous ride-hailing works — the operating data says it does. Waymo has logged more than 220 million autonomous miles and reports more than 94% fewer serious injury crashes than human drivers. It delivers around half a million paid rides a week. Uber, by comparison, operates in over 15,000 cities and completed 3.9 billion trips last quarter.
The harder question is what happens to the vehicles between rides.
Two hours of dead time per reset
A robotaxi that finishes a ride with a dirty interior or a low battery currently has one option: drive, empty, to a centralized depot. That trip can run up to 15 miles, and each reset costs roughly two hours of vehicle downtime.
The utilization cost shows up in the data. California figures show that of 86 million Waymo miles reported, only 54% carried a passenger. Not every empty mile is a depot run — but every depot run is an empty mile.
The grid clock runs in years
Depot infrastructure faces a second constraint: power. A new circuit for a charging site takes 1.9 years. A substation upgrade takes 2.8 years. A new substation takes 8.9 years. For an industry targeting thousands of cities, per-city timelines measured in years do not scale.
A fragmented ecosystem needs shared infrastructure
The market is also fragmenting. Waymo originally ran its full stack in-house, but even it now sources Ojai vans from Zeekr and depot operations from Moove, Avomo, Avis, and Flexdrive. Meanwhile, Wayve, Nuro, and Mobileye supply autonomous driving systems to Nissan, Lucid, Volkswagen, and Mercedes — the latter pairing with NVIDIA. The result is dozens of fleets running different vehicles, all needing charging, cleaning, and inspection several times a day.
Dan Keene, co-founder of Aseon Labs — a Y Combinator-backed company — previously co-founded Pushme, the battery-swapping network acquired by TIER. His company is building a robotic charging, cleaning, and inspection station that arrives on a flatbed and is designed to go live within a day. Investor attention is arriving too: Andreessen Horowitz raised $1.1 billion for the physical buildout of AI.
The cars can already drive themselves. They still cannot charge, clean, or inspect themselves. Somebody has to build the pit.
For manufacturing and operations engineers, the takeaway is familiar: the vehicle was never the hard part. The operations were.
Is this your company?
This article features your business. Claim it to add your logo, contact details, and a link to your website — or upgrade to reach more buyers.
Did you know 80% of Press Releases trigger AI content warnings? Reach out and the M4S team can assist.