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

# Why robotaxis stall between rides

Robotaxis have demonstrated driving ability: Waymo has logged more than 220 million autonomous miles, reports about 94% fewer serious injury crashes than human drivers, and gives around half a million paid rides a week. But driving is only part of the problem. The industry's practical bottleneck is what happens to the vehicle between rides: charging, cleaning, and inspection.

That operational gap is why a company can run robotaxis in a handful of cities but scaling to thousands becomes infeasible.

# Constraints that break city-by-city depots

Land scarcity and competition for powered parcels: industrial-zoned lots with sufficient power are in high demand. Data center developers and large cloud firms are outbidding prospective depot operators for the same sites.

Local variation: Cities differ in layout and climate. A depot design that works in Phoenix parking sprawl won't fit medieval street centers like Zurich or extreme climates like Riyadh. That leads to bespoke projects with higher cost and paperwork per city.

# Changing industry structure raises interoperability needs

The result will be many fleets of different vehicle types operating on multiple networks. That increases demand for servicing infrastructure that can support any vehicle and any operator without building a unique depot for each.

# What unlocking scale looks like

The article argues for a standardized, modular servicing unit — deployable quickly, configurable per city, and usable by any fleet or network — analogous to how the shipping container standardized ports. Such a unit would reduce site engineering, speed deployment to days rather than years, and fit in small industrial bays.

Aseon Labs (the author's company) is building a robotic station designed to be truck-delivered and operational within a day. The case for that product is arithmetic: the long per-city timelines for traditional depots cannot support thousands of cities.

# Practical implication for operators and planners

# Bottom line

Driving software is advancing fast and commercial deployments are expanding, but scaling robotaxis globally depends on solving the unglamorous, local problem of servicing between rides. Without a standardized, rapidly deployable servicing model, the industry will face land, power, and regulatory bottlenecks that stall growth.