How I check whether a VRP solution is actually feasible

A VRP route can look reasonable on a map and still be infeasible.

When I check a routing result, I do not start from the total cost alone. I start from the decoded structure: which vehicle serves which customer, in what order, under which constraints, and with what resource state.

My basic checklist:

• every required customer is served
• no customer is served twice unless the model explicitly allows it
• each route starts and ends at the correct depot
• vehicle capacity is not violated
• time windows or service-time logic are respected
• EV battery constraints are not silently ignored
• charging station visits are feasible under the energy logic
• conventional and electric vehicle costs are computed under the correct vehicle type
• the plotted route matches the decoded route
• the final result table matches the same solution

For mixed-fleet VRP, feasibility is vehicle-specific. A route that is feasible for a conventional vehicle may be infeasible for an electric vehicle because of range, charging, or energy-consumption logic.

This is why feasibility checking is not a small technical detail. It is the line between an optimization result and a decorative route figure.