Guide · Optimisation

Freight network optimisation: choosing the plant, carrier and truck for every depot at once.

When each depot is planned on its own, every lane looks sensible and the network as a whole still costs more than it needs to.

Freight network optimisation chooses, for every depot, which plant supplies it, which carrier moves it and which truck type is used, so that the total landed cost of the network is as low as possible within real limits.

What the model decides and respects

RuleIn plain words
DemandEvery depot gets its tonnage, or the shortfall is reported
Plant outputNo plant ships more than it makes that month
Fleet in truck-daysA truck on a long lane completes fewer trips than one on a short shuttle
Carrier shareNo carrier takes more than its cap of any depot
CommitmentsContracted minimum volumes are honoured
Whole trucksYou cannot dispatch two-thirds of a truck

What it minimises

The sum over all trips of landed cost per trip, where landed cost includes the minimum billable load, rate, fuel surcharge, toll and handling, plus a penalty for any unserved demand. The penalty means you always get a plan back, with the exact shortfall shown rather than hidden.

Getting started

  1. Collect the masters: plants, depots, lanes, carriers, fleet, rate cards and demand.
  2. Validate them row by row; fix or reject bad rows with a reason.
  3. Write the policies down once: carrier caps and commitments.
  4. Solve, then compare the plan with your usual lane-by-lane assignment.
  5. Price each policy: what does a carrier-share cap cost the network?

For truck fill within one shipment, see truck load planning; for carrier performance, see the carrier scorecard.

How FreightPlan does it

FreightPlan loads your masters (plants, depots, lanes, carriers, fleet, rate cards and demand) from CSV, validating every row: a bad row is rejected with the reason and the fix, and the rest of the file loads. It prices every option with one landed-cost formula (billable load × rate × (1 + fuel surcharge) + toll + handling) and uses a Google OR-Tools optimiser to choose the plant, carrier and truck type for every depot, in whole trucks, within 1% of optimal, subject to six written rules: demand, plant output, fleet truck-days, carrier share, contracted commitments and whole trucks. Unserved demand is reported, never hidden. At month-end the same formula prices the trip log, the gap is split into volume and rate so the two add up exactly to the total, and the rate gap is split into five causes (empty truck space, rate card, unplanned trips, detention and service), each with an owner and an action.

Questions

What is freight network optimisation?

Choosing the plant, carrier and truck type for every depot together so the network’s total landed cost is lowest within demand, plant output, fleet, carrier-share, commitment and whole-truck limits.

Why plan in whole trucks?

Because a part-filled truck still costs a trip or the minimum billable load; fractional plans understate real cost.

What if demand cannot be met?

A good model still returns a plan and reports the exact shortfall, instead of failing or hiding it.

See it on your own data. FreightPlan — Budget freight, explain the gap. Book a 30-minute working session with an engineer.

General guidance, current as of the date above. Figures and examples are illustrative unless a source is linked.