Public transport is full of moments that shape how people feel about the cities they live in. A smooth journey builds confidence. A single failure can derail a day. The truth is, many of the problems that frustrate passengers are avoidable.
And that is where AI is starting to make a meaningful difference.
At Transport Ticketing Global 2026, I had the opportunity to sit down with Steve Ramsay from Innovatious for a fireside chat about how our industry can move beyond abstract conversations about AI and start using it to solve real operational problems.
Not “future potential” or “what might be possible”, but the practical applications that are already reducing breakdowns, improving reliability, protecting revenue, and making daily travel easier for millions of people.
The shift we’re seeing is simple but profound. AI is no longer a technology story. It’s an experience story. And when it’s done well, passengers never even notice it’s there.
Why this matters for passengers
I opened with a simple scenario. You’re rushing to work or heading to an interview. You reach the station, and the gates are down. You miss your train. Your day derails. These moments are small individually, but collectively they shape how people feel about public transport.
We now have the tools to prevent them.
Predictive maintenance that fixes problems before they happen
Through our research partnership with Imperial College London, Cubic is using AI to predict when ticket gate equipment is likely to fail. By training data analysis models on historical maintenance records and real‑time equipment data, engineers receive alerts before passengers are affected. They can arrive with the right parts, complete fewer site visits, and prevent breakdowns that cause queues and stress.
This isn’t theoretical. It’s live. It’s working today in London, Los Angeles, and Chicago.
Tackling fare evasion with computer vision that protects privacy
UK agencies estimate £240 million lost annually to fare evasion, while TfL alone loses £130 million. Revenue protection teams at TfL clearly recognised the scale of the challenge.
Our FEnX gate, powered by the FLARE computer vision platform, analyses human behaviour in real time to spot unusual patterns. It does this without biometrics. No facial recognition. No personal identifiers. The system simply classifies general attributes, such as “adult male, average height”, and looks for behaviour that stands out.
One example from New York resonated strongly with the audience. During our pilot of 40 FEnX gates across seven stations, the system detected a person stationed on the paid side of the gate opening it repeatedly to let others through while charging them 50 cents. Enforcement teams were guided directly to the exact location within minutes.
Beyond evasion: identifying errors, misuse and operational trends
AI helps agencies see patterns at scale:
- multiple taps on the same concession card
- “forgot to tap out” events
- unusual behaviour that may signal account misuse
- hotspot maps highlighting where evasion is occurring
- busy‑station trends that support better staffing and fare policy decisions
This is where AI becomes truly powerful — not just in detecting problems, but in helping agencies shape smarter networks.
Built on responsible AI principles
Public trust matters. The transport industry is heavily regulated, and rightly so. Our systems use only the data required, comply with all local legislation, and embed responsible AI by design. FEnX and FLARE operate within strict boundaries that protect passenger identity while still enabling meaningful insights.
Where theory meets reality
One of the best questions came from the audience. I was asked whether predictive maintenance is truly deployed live today, and whether it relies on real‑time or historical data. That allowed me to explain how our models use both historical records to understand long‑term patterns and real‑time data to dynamically refine predictions.
A final thought: AI should be invisible
AI is real. But it is more real when it makes travel effortless for everyone. When it creates no noise and no headlines. When it simply works in the background, so people can get where they need to go.
We are building a future where AI quietly solves genuine operational problems, supports frontline teams, and makes journeys seamless for millions.