Last week, during testing of one of the three Siemens trains returned to service by Connex, the following incident happened, revealing an underlying problem with the basic safety system on the train that no-one wants to touch with the proverbial barge pole.

To understand the implications of what occurred you need to know this: almost every signal in the metropolitan network is equipped with a “train stop”. This does what its name implies. If the signal is at red (also termed “stop” or “danger”), a lever is raised above rail level. This lever interacts with a device on a suburban train called a “trip cock”.

When the trip cock interacts with a stop signal, the emergency brake is activated and the driver’s ability to operate the motors is removed. It’s an obvious failsafe system.

Last week, during the test, the “safety apparatus” of a Siemens failed in an unsafe way. The set had been skidding and, as a consequence, passed a signal at the stop position. No fault of the driver – the train had skidded for 1.5 km. The driver had the brake control in the emergency position as part of the test.

When the train interacted with the signalling system, the trip cock was activated. But it immediately “reset”, releasing the emergency brake, despite the fact that the train was still at high speed AND THE BRAKE CONTROL WAS STILL IN THE EMERGENCY POSITION.

I could get even more technical, but the effect is that the basic principle of rail operations – that the train stops when there’s a disruption to its safety systems (and/or the driver is incapacitated) – seems to have been disabled by the same software modifications designed to “fix” the spate of low speed overshoots before the last state election. Ask your friends at Leader Newspapers about their photo of the incident at Williamstown.

In effect, it means that there is no effective “deadman’s” device on the Siemens trains. Or at minimum, the train prematurely returned to service. (Siemens seems to do a separate software fix for every faulty train, so it “might” only be the one.)

This arises, purely and simply, from the fact that the trains are not, and never were, fit for the safe carriage of passengers in Melbourne. Siemens, which has the maintenance contract for the trains, is keen to minimise its costs, and has apparently designed the trains to protect the expensive motors. In any event where the motors might be at risk, control of the emergency brake, and, in fact, the normal brake, is removed from the driver.

As reported in The Age a couple of weeks ago, this leaves the driver no option but to use the parking brake. But the parking brake is designed to hold, not stop the train. Just like that on your average car. The Siemens parking brake will partly release if the train is in motion – to prevent a skid. So it’s not a reliable brake, and certainly not an emergency brake.

Bottom line: the Siemens fleet should be off the rails until the fault is fixed – and it’s not fixed yet. Again, avoiding technicals, the fact that the problem is alleged not to happen on six carriage trains points to a fault in the train’s basic safety systems. They keep fiddling with software without a plan or understanding of the complex physical mechanics of the machine.

The Siemens trains suffer from a basic design flaw that can only be fixed by an expensive re-engineering of the safety system. Not just software, but real pipework and welding. At present, if a driver is incapacitated under certain conditions, the train will not stop when it interacts with the signalling system.

There’s a political quick fix on the front burner which fixes nothing.