Fare evasion costs UK rail £240 million a year. New research reveals how data can help safeguard revenue.

“Short Ticketing”, where passengers buy a ticket for part of a journey instead of end to end, is hard to uncover and even harder to prevent.

Cubic Transportation Systems, working with Imperial College London’s AIDA Lab (via Imperial Consultants), published a new whitepaper showing how advanced data techniques can tackle this hidden problem.

The report dives into a smart detection framework built on 6.5 million journey records from 100 stations over seven days, revealing how data can help tackle the problem head-on.

Cracking down on “Short Ticketing”

Using a mix of machine learning models, the research team analysed travel patterns to find anomalies that suggest short ticketing. The results are impressive: they identified 30 high-risk stations and uncovered five distinct fraud patterns.

Why does this matter? Because it gives operators the tools to act. Enforcement can focus on the places and behaviours most likely to involve fraud. That means better protection for revenue and a fairer system for honest passengers.

Using data and technology to protect rail revenue

Using our combined analysis, we spotted five patterns of short ticketing:

• Ghost Station: Stations where passengers only enter or exit but never complete full journeys.
• Black-Hole Station: There are a disproportionate exit-only ticket patterns at these stations.
• Fake-Origin: Passengers may be using these stations to simulate a legitimate journey start but boarding elsewhere.
• Function-Loss: Extreme declared destination/declared origin ratios at these stations indicate a breakdown in normal usage.
• Micro-Trap behaviours: Subtle anomalies shown in these low-volume stations are harder to detect but may signal emerging hotspots for exploitation.

Finding these patterns matters. It helps to plan smarter revenue protection – from targeted inspections to better gate logic – tailored to each station.

We also learned which stations were most likely to show certain types of short ticketing. For example, one airport station had a strong ghost station pattern, with 63% of tickets being entry-only and 37% exit-only. A busy city centre station showed a black-hole pattern, with 62% exit-only tickets and a 24% imbalance between entries and exits.

These insights mean agencies can send inspectors and revenue protection teams to the right places, instead of carrying out costly blanket checks. This targeted approach could recover millions in lost revenue. And because the system uses anonymised operational data, passenger privacy is protected while giving agencies practical, actionable information.

To find out more, you can read the full whitepapers on Cubic’s website or visit the AIDA Lab at Imperial College London.