Autonomy at Scale

vehicles · depots · power · people

20,000 robotaxis need about a gigawatt-hour of charging a day. At scale, autonomy is a depot and power problem.

Fleet plans: Waymo had more than 4,000 vehicles in September 2026 and plans about 8,000 by the end of 2026 and 20,000 by the end of 2027 (Semafor); Aurora targets more than 30,000 driverless trucks by 2030. ElectronsX estimates for 20,000 robotaxis: about 1 GWh of charging a day, about 40 depots of 500 vehicles, and on the order of 1,000 remote-assistance staff. FLEET PLANS (VEHICLES) Waymo, Sept 2026 4,000+ Waymo plan, end-2026 ~8,000 Waymo plan, end-2027 20,000 Aurora target, 2030 30,000+ WHAT 20,000 ROBOTAXIS NEED (ESTIMATES) ~1 GWh/day charging energy ~40 depots of 500 vehicles ~1,000 staff remote assistance
Waymo now
4,000+
Waymo end-2027
20,000
Energy per robotaxi
~50 kWh/day
Depot size
~500
Hub
Autonomy

Finding

Past a few thousand vehicles, autonomy stops being a software problem and becomes a vehicle, depot and power problem

Waymo's plan to grow from about 4,000 to 20,000 robotaxis by the end of 2027 is a manufacturing and real-estate program as much as a software one. At roughly 200 miles a day and 0.25 kWh a mile, 20,000 robotaxis draw about 1 GWh of electricity a day; parked in depots of about 500 vehicles, that is around 40 sites, each needing several megawatts of grid service, cleaning bays and inspection lanes.

The other constraints scale the same way. Every vehicle needs a sensor kit and an inference computer, which ties fleet growth to lidar output, automotive chip supply and, for Tesla, its own AI5 ramp. Remote assistance - people who help vehicles through situations they cannot resolve - grows with the fleet even as the ratio of vehicles per person improves. And each new city needs mapping, approvals and local depot partners. The fleets that scale fastest will be the ones that solve these physical constraints, which is why the depot and grid pages sit in this hub.

The estimates above are ElectronsX planning figures, not operator disclosures. See Autonomous Fleet Depots and Fleet Energy Depots.

Constraints

What limits fleet growth

Autonomy scaling constraints
ConstraintWhat it coversLimiting factor
VehiclesPurpose-built robotaxis (Cybercab, Zoox, Zeekr-built Ojai) and retrofit linesFactory output and retrofit capacity
SensorsCameras, radar, lidar per vehicleLidar and camera module supply
Inference computeOne or two autonomy computers per vehicleAutomotive-grade AI chips (Nvidia, Tesla AI5, Horizon)
DepotsCharging, cleaning, inspection, stagingSites near demand; grid service
Power~50 kWh per robotaxi per day (estimate)Utility interconnection timelines
Remote assistanceStaff who resolve edge casesHiring and tooling; ratio improves with software
ApprovalsCity and state permits; NHTSA exemptions for vehicles without controlsZoox exemption capped at 2,500 vehicles a year

Sources

Sources and scope

Scope
Physical limits on scaling autonomous fleets
Sources
Semafor via Reuters (October 2026); Aurora 2026 Investor Day; NHTSA (Zoox exemption, July 2026); ElectronsX estimates
Updated
October 2026