Model the operation before choosing hardware
A successful swapping network is an operations system, not only a battery and cabinet purchase. Route length, vehicle efficiency, shift patterns, rider behavior, electricity availability and service response determine how many charged packs must be available at each location. Start with the busiest hour and the worst normal operating day.
Choose a battery around the vehicle duty cycle
Match voltage, current capability, enclosure dimensions, connector, communication protocol and mounting method to the vehicle. Then compare LFP and NMC chemistry. LFP can support long-cycle, thermally stable operation; NMC can reduce weight or increase energy in a similar envelope. The correct choice depends on route range, handling limits, climate and lifecycle economics.
Validate the pack and vehicle as a system. Nominal voltage alone does not confirm compatibility. Motor controller limits, regenerative braking, discharge peaks, connector temperature and BMS communication all need testing.
Estimate the number of packs
Each vehicle needs an installed pack, plus a charging pool and an operating reserve. A simple starting model uses daily energy demand divided by usable energy per pack, then maps when packs return to each station. Add reserve for uneven demand, maintenance, delayed returns and packs approaching end of life.
Avoid applying one fleet-wide reserve percentage without simulation. A distributed network with several small stations behaves differently from a depot where every vehicle returns at the same time.
Size cabinet throughput, not only slot count
A 12-slot cabinet does not automatically deliver 12 fully charged packs whenever riders arrive. Charging power per slot, total site input, starting state of charge and dwell time determine throughput. Model each pack as it enters, charges and leaves during the peak window.
The site electrical connection may be the limiting factor. Smart charging should prioritize packs by departure need while keeping the cabinet below its total input limit. Tariffs may also favor charging outside the local peak period.
Design the rider and service workflow
A fast exchange still needs identity, payment or fleet authorization, battery health checks and exception handling. Define what happens when a connector is damaged, a pack is too hot, a slot is offline or a rider returns the wrong asset. The cabinet, cloud platform and local service team should use the same battery identity and status records.
- Rider authentication and swap authorization
- Battery serial number and lifecycle tracking
- Remote alarm and cabinet status monitoring
- Local spare-parts and escalation process
- Safe handling and damaged-pack isolation
Pilot before scaling
Run a controlled pilot across representative routes and weather conditions. Measure energy per kilometer, swaps per vehicle, queue time, cabinet utilization, charging downtime, pack temperature and service events. Use the results to update pack quantity, station placement and charging rules before ordering the full network.
For a supplier discussion, prepare vehicle specifications, fleet size, shift schedule, route data, electricity supply, target locations and required software interfaces. That turns a general product inquiry into a system proposal that can be tested against real operating targets.