Perishable product distribution requires efficient route planning because delivery duration, route sequence, and transportation conditions directly affect product quality, service reliability, and operational performance. This study evaluates hard and penalty-based soft time-window routing for crystal-ice distribution in a small-scale enterprise distribution system comprising one depot, 60 scheduled customers, and two non-refrigerated vehicles. A quantitative computational case-study approach was applied using the Vehicle Routing Problem with Time Windows (VRPTW) and the Cheapest Insertion Heuristic (CIH). Soft time-window performance was evaluated using lateness penalties of λ = 0.1, 0.5, 1, 2, and 5 km/min. The hard time-window solution produced a fleet distance of 86.87 km and zero lateness. At λ = 0.1, the soft model reduced the fleet distance to 85.29 km but generated 8.95 minutes of total lateness across two customers. Increasing the penalty to λ = 0.5 eliminated all lateness and produced the same 86.87-km route as the hard model. Compared with the observed company baseline, the selected zero-lateness route reduced fleet distance from 99.84 to 86.87 km (12.99%) and route exposure duration from approximately 1,200 to 1,043.59 minutes (13.03%). Model-estimated product loss decreased from an observed baseline of 580 kg to 507.95 kg, corresponding to an estimated reduction of 72.05 kg (12.42%). These findings show that soft time-window penalties can reveal the trade-off between marginal distance savings and service punctuality.
Copyrights © 2026