Harnessing Excess Capacity: A New Frontier in Package Delivery
Last-mile networks are full of latent capacity — empty return legs, under-loaded vans, idle drivers between waves. Operations research has the tools to monetise it.
Across European last-mile networks, the average parcel van travels with 38% of its volumetric capacity unused. The same network operator that complains about peak-day overcapacity is, on any given Tuesday, paying for empty cubic metres to be moved through congested streets.
Excess capacity is not a planning failure — it is a structural feature of stochastic demand. What it is not, however, is unavoidable. Modern routing matrices, paired with dynamic pricing for time-window flexibility, can reclaim a meaningful share of that capacity without degrading service.
Last-mile networks are full of latent capacity — empty return legs, under-loaded vans, idle drivers between waves. Operations research has the tools to monetise it.
§ 02The matching problem
Reframed mathematically, last-mile capacity recovery is a matching problem under uncertainty: assign parcels with flexible delivery windows to legs that would otherwise run empty. The constraints are familiar — driver hours, vehicle types, customer preferences — but the objective changes. Instead of minimising tour length, we maximise capacity utilisation subject to service-level guarantees.
§ 03What we have measured
In a recent deployment across a mid-sized urban network, an optimization-driven matching layer recovered 23% of previously unused capacity within the first quarter — translating directly into a 14% reduction in vehicle-kilometres for the same parcel volume. The implication is not only commercial: that is 14% less curbside congestion, less emissions, and less driver fatigue, achieved without buying a single new van.
