Rail logistics / selected experience
Match daily capacity to demand
A large rail logistics operator struggled to predict how many trains each terminal would need.
Rail logistics
From the problem to a practical change.
- Terminal demand
- Two-day forecast
- Train allocation
What changed
For most destinations, the reported average prediction error was fewer than two train allocations.
Put the result in context
The result applies to most destinations, not every route. The test period and dataset size are not stated.
How it worked
Daily train-allocation forecasts with a two-day planning horizon.
Single-variable and multi-variable models were compared, with forecasts shown in a practical planning view.
Previous project experience from our delivery team’s portfolio. Outcomes are reported in the project portfolio and have not been independently audited by WitFlow. They are not forecasts for a new project.
Could this help with your problem?
Tell us what is happening in your business. We’ll look at where a similar approach could help, and what would need testing first.
