Finding
In a mixed depot, robotaxis and trucks set the power peak, robots and drones set the number of charge points, and the battery is sized to the overlap, not to daily energy
In an illustrative depot with 100 robotaxis, 40 vans, 200 humanoids and 30 drones, the robotaxis hold two-thirds of the 6.8 MW of connected charging but only a tenth of the 300 charge points; the humanoids hold two-thirds of the points and 3% of the power. The site uses about 10 MWh a day, an average of 0.43 MW, but the evening robotaxi return on top of vans starting their overnight charge peaks near 4 MW.
A 2.5 MW utility service with a battery of about 1.5 MW and 3 MWh covers that two-hour overlap; sizing the service to the peak instead would mean a larger transformer and a longer interconnection queue for capacity used a few hours a day. The same arithmetic explains why humanoid and drone docks hardly move the electrical design but dominate the floor plan, cabling and network ports: 230 small connections that each need a dock, a data link and a slot in the scheduler.
Factories already run this way. Giga Texas charges employee cars, Semi trucks and yard tractors, runs autonomous mobile robots and forklifts, and is training Optimus humanoids on one campus - see the AustinIO page on Tesla Giga Texas. BYD and CATL campuses in China mix electric buses, logistics trucks and AGVs the same way. The operating side of a shared yard - staging, dispatch and maintenance - is covered on Energy Autonomy Yards.
Worked example
Sizing an illustrative mixed depot
ElectronsX illustrative calculation; assumptions in the rows.
| Asset | Fleet | Charge points | Connected power | Daily energy | Evening peak load |
|---|---|---|---|---|---|
| Robotaxis | 100 at ~50 kWh/day | 30 DC at 150 kW | 4.5 MW | 5.0 MWh | 3.0 MW (20 stalls busy) |
| Delivery vans | 40 at ~80 kWh/day | 40 DC at 50 kW | 2.0 MW | 3.2 MWh | 0.8 MW (managed) |
| Humanoids | 200 at ~7 kWh/day | 200 docks at 1 kW | 0.2 MW | 1.4 MWh | 0.15 MW |
| Drones | 30 at ~20 kWh/day | 30 pads at 3 kW | 0.09 MW | 0.6 MWh | 0.05 MW |
| Site | 370 assets | 300 | 6.8 MW | 10.2 MWh (0.43 MW average) | ~4.0 MW against a 2.5 MW service |
Power envelope
Power envelope by asset class
| Asset | Battery | Charging power | Pattern | Scheduling priority |
|---|---|---|---|---|
| Fleet EVs and vans | 60-120 kWh (vans); 400-900 kWh (trucks) | 19 kW AC to 350 kW DC; 1 MW+ for trucks | Overnight or between shifts | Flexible within the dwell window |
| Robotaxis | 50-100 kWh | 150-350 kW DC, or inductive pads | Short top-ups across long service hours | High: idle vehicles lose revenue |
| Humanoid robots | About 1-3 kWh | 0.5-5 kW docks | Frequent dock returns | Low per unit, but hundreds of units |
| Drones | 0.1-2 kWh | 0.5-10 kW pads, or battery swap | Bursts tied to missions | Mission-timed; swap avoids waiting |
Battery roles
What the depot battery does
Size it to coincident power and peak duration, then check that it still covers critical loads in an outage.
| Function | Role | Why it matters |
|---|---|---|
| Peak shaving | Covers overlapping charging events | Smaller utility service and lower demand charges |
| Load shifting | Stores cheap or solar energy for later | Lower energy cost; matches charging to tariff windows |
| Power quality | Buffers fast load ramps | Stable voltage with many chargers switching |
| Resilience | Keeps priority assets charging in an outage | Continuity for autonomous and emergency fleets |
Sources
Sources and scope
- Scope
- Multi-Fleet Depots
- Sources
- ElectronsX illustrative calculation; manufacturer battery specifications
- Updated
- October 2026