A depot with 300 vehicles and 40 stalls cannot charge every vehicle simultaneously; it must cycle vehicles through the stalls overnight while guaranteeing every route-critical vehicle reaches its required state of charge by its departure time. This is the canonical problem that fleet-scheduling software is built to solve.
The overnight window from 10 p.m. to 5 a.m. is discretized into twenty-eight fifteen-minute slots across forty stalls, yielding 1,120 stall-slots to allocate among 300 vehicles each requiring between two and six slots of charging. The solver's objective function minimizes the count of vehicles that would miss their required departure state of charge, subject to the constraint that no fifteen-minute slot exceeds the depot's contracted grid connection capacity.
When the aggregate demand of all vehicles wanting to charge simultaneously exceeds the depot's grid connection ceiling, the solver does not fail; it staggers session start times and applies partial-power charging to lower-priority vehicles, trading charge rate for schedule feasibility. Route-critical vehicles are pinned to full-power slots first, and the remaining power budget is distributed among the rest by descending priority score.