Design an EV Battery Swapping Network

Problem Statement: EV Battery Swapping

A battery swapping network lets an electric vehicle exchange a depleted pack for a fully charged one in a few minutes — the refuel-speed answer to slow charging. The system's job is not charging a car; it is managing a fleet of fungible battery packs as moving inventory across stations, guaranteeing that a charged, healthy, compatible pack is waiting when a vehicle arrives.

The defining difference from EV charging is that the battery is a tracked asset that circulates: a pack lives in a vehicle, comes out depleted at a station, charges, and goes back out in a different vehicle. Each pack carries its own state-of-health (SoH) — capacity fade and cycle count from age and use — and a degraded pack must never be handed to a customer. So the unit of the system is the pack, individually identified and tracked through its lifecycle, not the kilowatt-hour.

A swap is a two-sided atomic operation: the vehicle docks its depleted pack into an empty slot and receives a ready pack, both in one transaction. That makes a station a small warehouse of packs in different states — charging, ready, in a vehicle, or quarantined for low health — and the central problem is keeping enough ready packs available for demand while charging the depleted ones cost-effectively.

The scale that matters is inventory and scheduling, not raw QPS. A network might run ~2,000 stations × ~20 slots holding ~1,000,000 packs in circulation, serving ~500,000 swaps/day (~6/s average, peaking at commute). The hard questions are: which pack do I hand out (SoH + compatibility), can I promise a ready pack at a reserved time (reservation against charge schedule), and when do I charge depleted packs to be ready without spiking the grid (grid-aware scheduling). It is a logistics-and-scheduling system wearing an energy costume.