Open Up Public Transport Performance Data

Anyone who expects a public transport system to always run on time and without faults or other problems is almost certainly kidding themselves. Like any man-made system, things just go wrong sometimes.

The sign of a good public transport system is how well the situation is handled when things do go wrong. Good public transport systems should not really be noticed on a day-to-day basis – they should just work.

It would seem quite reasonable to offer incentives for companies running the systems to do a good job – theoretically at least, it eliminates complacency and gives their workers a good reason to do the very best that they can for the paying public.

In Victoria, the current operator of Melbourne’s Metro Trains is the MTM Consortium. They are rewarded for meeting punctuality and reliability targets, and penalised if they fail to meet them.

No problem there – unless of course, they are gaming the system?

“Metro is terminating trains part-way along the line hundreds of times a week, dumping passengers short of their destination in an effort to meet lucrative government-mandated performance targets.”

As I understand it, a service that doesn’t run the full length of the line is not considered to be a scheduled service, so it isn’t counted in the statistics covering “scheduled services”.

For example, a 4:05pm train that was supposed to travel from the City to Lilydale is suddenly terminated at Ringwood because it was running a bit late, and becomes an “unscheduled service”, because there is no “4:05pm City to Ringwood” service on the timetable. These “short running” trains don’t seem to affect the result in any way.

That’s pretty handy for when you’re trying to meet reliability and punctuality performance targets on the promise of getting money for meeting them – and it’s been pretty lucrative for them too:

“Metro has consistently met its targets in recent years. In the 18 months to September, it received $16.7 million in incentive bonuses. But the good results have been achieved in part by a readiness to alter services to stick to the timetable.”


“In all, 550 services were altered or did not run in the first week of March, although just 151 services were officially recorded as cancelled, putting Metro narrowly inside the 98 per cent reliability threshold that triggers customer compensation.”

They landed just inside the threshold huh? Gee, that was fortunate!

But where does the data about the performance of the network come from? It used to be manual, but apparently that has changed:

“PRS uses track sensors that automatically record train arrival and departure times at stations, reducing the government’s reliance on data provided by Metro to inspect whether it has met targets that can secure it million of dollars in quarterly bonuses.”

So theoretically the government has the data, yet Metro are gaming the system and still getting away with it?

What’s the solution? Become transparent and release the raw data publicly – let’s see it.

The problem with this whole compensation system is – (as we can see) – that we don’t know that all of the late/changed/cancelled services are being included in the data.

It seems relatively clear that they are not.

When you get off a train at your station and it has arrived 20 minutes late, how do you know it will be included in the next round of data as having been so late?

If we had the data – openly and transparently – we could all go check, and find out whether or not these performance bonuses are justified or not.

Some accountability, if you’d like to word it like that.

I want to be able to get off a late running train, and be able to go to a website – (within a reasonable time frame) – and confirm that the fact this train ran late – (or ran short, or was cancelled) – is in the data.

At the moment it seems like we as taxpayers are being ripped off, and the government knows it.

I’ve suggested this kind of transparency before, but I was told by V/Line at the time that it was almost impossible to provide this data. It was even inferred that the data didn’t even exist to be able to be published.

So what data were the reports being produced from? Exactly.

How about it government? Want to do the right thing?