Your train arrived on time. So why is the rail network disrupted?

On-time performance is the headline metric for most rail networks. It's simple, visible, and easy to report. But it only tells you when a train arrived or departed, not how it got there, or what it disrupted along the way.

And that missing detail matters, because many of the problems that cause delays across the network don't show up in station arrival data at all.

What on-time data doesn't show

Most operational reporting is built around stations. If a train arrives and departs within tolerance, it's recorded as "on time".

But between those stations, a lot can go wrong.

For example:

• A train runs slightly under speed for several miles due to a technical issue
• A driver takes longer than planned at a signal due to unclear routing
• A train dwells just long enough at multiple stations to stay within thresholds

Individually, these issues may not trigger an alert or breach a KPI.

But collectively, they can:

• Delay following services
• Reduce available headway on busy sections
• Increase knock-on reactionary delay elsewhere on the network

By the time a delay is recorded, the root cause is often several miles (and several minutes) away. These hidden delays are sub-threshold delays.

The hidden cost of "sub-threshold" delays

A train may arrive on time but if it:

• occupies a section longer than planned
• forces other services to slow or wait
• creates uneven spacing behind it

…it can increase overall delay minutes without ever being labelled as the problem.

Over time, these small issues add up. The results are very real:

• More compensation claims from passengers
• Higher operating costs from recovery actions
• Performance penalties from regulators
• Reduced trust, leading to lower future ticket sales

None of these show up in a simple on-time arrival chart.

Seeing the whole journey, not just the timestamps

Better performance comes from understanding what happens between stations.

Emu Analytics' software shows rail operators:

• Real-time train location across the network
• How long trains occupy sections, not just stations
• Where speed reductions, stop-start movement, or abnormal behaviour occur
• Patterns that repeat on specific routes, times, or rolling stock

This makes it possible to:

• Intervene earlier, before delays propagate
• Target fixes to the sections that cause the most reactionary delay
• Adjust plans using evidence, not assumptions

In aviation, this approach has reduced turnaround delays and improved schedule reliability by focusing on where time is actually lost. In rail, it's already helping teams identify the real causes behind "mystery" delays and take more effective action.

On time is necessary, but it's not enough

On-time arrival will always matter. Passengers care about it, and so do operators.

But a railway that focuses only on station timestamps risks missing the problems that quietly undermine performance every day.

Excellent performance comes from understanding the full journey (including the parts that don't show up in traditional reports) and fixing issues before they affect everyone else on the network.