Real-Time Train Service Matching

Train operating companies have to deal with a major safety issue—how often should they maintain their trains? Maintenance periods largely depend on the distance a rolling stock has travelled, but due to disruptions, companies can lose track of which service is run by which rolling stock.

We are helping them solve this problem with an algorithm that uses existing data streams to match services to rolling stock in real-time.

Real-time processing

Runs in real-time and produces results incrementally and efficiently, altering only the part of the result that is affected by the changes.

Consolidates 3 data sources

Combines 2 real-time data streams and daily allocations data to determine the most likely matchings of the day.

Visualisation of results

Allows user to see the matchings and provides the data used to make a decision, so they can analyse the matchings.

Search for matchings

Search for services and units of interest, or find all the cases where the algorithm disagrees with the data given.