Skip to main content

Logistics Depot Location

Capacitated Facility Location (CFLP) — Lagrangian Relaxation

A logistics company must decide which of 5 candidate depot sites to open and what capacity to provision. Each depot costs €1–3M to establish. Transport from depot to customer zones adds €0.5–2.0 per parcel-km. Wrong locations lock in 10-year excess delivery costs.

Where This Decision Fits

Logistics planning chain — the highlighted step is what this page optimizes

Network DesignDepot location
Fleet PlanningVehicle sizing
RoutingVRP / TSP
Crew SchedulingDriver rostering
Dynamic OpsReal-time dispatch

The Problem

Minimize total fixed plus transport cost across candidate depots

A logistics company evaluates 5 candidate depot locations to serve 10 demand zones. Each depot has a fixed opening cost (ranging from €1M to €3M), a maximum parcel-handling capacity, and a geographic position. Each demand zone requires a specific daily parcel volume, and the cost to transport parcels from depot i to zone j depends on distance and a per-parcel-km rate.

The objective is to select which depots to open and assign each zone’s demand to opened depots without exceeding capacity, minimizing total fixed cost plus transport cost. This is the classical Capacitated Facility Location Problem (CFLP), proven NP-hard.

CFLP Formulation minimize   Σi fi · yi + ΣiΣj cij · xij   // fixed + transport cost
subject to
  Σi xij = dj   ∀ j    // each zone's demand fully met
  Σj xij ≤ Qi · yi   ∀ i    // depot capacity if opened
  yi ∈ {0, 1}   // open or not
  xij ≥ 0   // parcels from depot i to zone j

Where fi is the fixed cost of depot i, yi is the binary open/close decision, Qi is the capacity of depot i, cij is the per-parcel transport cost from depot i to zone j, dj is the demand of zone j, and xij is the parcel flow variable.

See Integer Programming theory

Try It Yourself

Select depots and assign demand zones to minimize total cost

Depot Location Optimizer

5 Depots · 10 Zones
DepotCapacity (parcels/day)Fixed Cost (€M)
ZoneDemand (parcels/day)

Ready. Click “Solve & Compare All Algorithms” to run.

AlgorithmTotal (€M)Depots OpenCapacity %Fixed (€M)Transport (€M)Time
Click Solve & Compare All Algorithms
Assignment details will appear here after solving.

The Algorithms

Three approaches to capacitated facility location

Constructive

Greedy (Cheapest per Demand)

O(n²m)  |  No approximation guarantee

Start with no depots open. At each step, evaluate the marginal cost-per-parcel of opening each closed depot: its fixed cost amortized over the demand it would capture, plus the transport savings from reassigning zones. Open the depot with the best ratio. Continue until all demand is covered or no improvement remains. Fast and intuitive but can miss globally optimal configurations.

Dual Bound

Lagrangian Relaxation (5 iterations)

O(K · n · m)  |  Provides lower bound

Relax the demand-satisfaction constraints into the objective using Lagrange multipliers λj. The relaxed problem decomposes into independent per-depot subproblems: open depot i if its reduced fixed cost minus the sum of profitable adjusted transport margins is negative. Update multipliers via subgradient ascent over 5 iterations. The best primal-feasible solution encountered gives an upper bound; the Lagrangian value gives a lower bound.

Improvement

Local Search

O(n · m · I)  |  No guarantee

Start from the Greedy solution. At each iteration, consider three neighborhood moves: (1) swap — close one depot and open another; (2) add — open an additional depot; (3) drop — close an existing depot if capacity allows. Accept the best improving move. Repeat until no single move improves the objective. Typically refines the greedy solution by 5–15%.

Real-World Complexity

Why practical depot location goes beyond the basic CFLP model

Beyond Capacitated Location

  • Multi-echelon networks — Hub-and-spoke designs with regional hubs, local depots, and last-mile micro-hubs add nested location decisions
  • Time-window constraints — Same-day and next-day delivery promises impose maximum distance or travel-time limits on depot-to-zone assignments
  • Demand uncertainty — Seasonal peaks (holiday surges) and stochastic daily volumes require robust or chance-constrained formulations
  • Multi-period expansion — Leases span 5–10 years; phased rollout must balance upfront investment with demand growth trajectories
  • Labor market access — Depot viability depends on local wage levels, driver availability, and workforce commuting patterns
  • Zoning and permits — Industrial zoning, noise restrictions, and truck-route limitations constrain feasible locations
  • Returns and reverse logistics — Depots that also handle returns have asymmetric flow, requiring additional capacity and sorting infrastructure

Key References

Foundational works in capacitated facility location

  • Cornuéjols, G., Fisher, M.L. & Nemhauser, G.L. (1977). “Location of Bank Accounts to Optimize Float: An Analytic Study of Exact and Approximate Algorithms.” Management Science, 23(8), 789–810.
  • Li, S. (2013). “A 1.488-Approximation Algorithm for the Uncapacitated Facility Location Problem.” Information and Computation, 222, 45–57.

Need to optimize your logistics network?

From depot location to fleet planning and last-mile routing, mathematical modeling can transform supply chain investment decisions. Let’s discuss how Operations Research can work for you.

Disclaimer
Data shown is illustrative. Depot names, capacities, costs, and zone demands are representative scenarios for educational purposes and do not reflect any specific logistics network or planning process.
Back
ESC