Technology Saves Blog From EU Debate

Yup! This week I am avoiding further commentary on the increasingly bizarre EU debate and turning my attention to driver-less cars.

Apparently the UK government has created an Intelligent Mobility Fund which is stumping up £100M to enable UK to grab 10% of the emerging driver-less car market.  Given that one of the players is Google, which currently has $178 Billion in cash one wonders whether this is money well spent.  But my question today is more fundamental – why are we seeking to develop autonomous road vehicles when we have yet to fully automate railways?

Of the four main modes of transport, road, sea, air and rail, by far and away the most complicated is road.  Air and sea have both had autopilots for decades, albeit ones still supervised by humans (although some military drones such as Global Hawk are completely autonomous). This is sort of understandable as if something goes wrong both are a long way from a fitter and, in the case of a plane, it is not possible to stop and wait for one to show up.  So ships and planes have reversionart modes in which humans take control.

Rail transport is simple.  A train can only go forwards, backwards or stop.  Changing lanes happens rarely, and the system for controlling that is separate from the system operating the train.  At the moment the train driver obeys signals, moves at the appropriate speed for the track and stops at stations.  The train is actually equipped with a “dead man’s pedal” to stop the train if the driver becomes incapacitated.

There is an easy, public measure of how safe train drivers are, and that is the Signals Passed at Danger (SPAD) rate.  This is when a train passes a signal instructing it to stop.  It can occur for a number of reasons and they are ranked by severity driven by the overshoot distance.  There are multiple causes, from driver missing signal completely, being on his (personal) mobile phone to misjudging the braking distance due to track condition (i.e wet, ice, leaves etc.)  This data is collected by the rail regulator (ORR) and published quarterly.  I phoned them and the fantastic Lyndsey Melbourne emailed me the SPAD data within 5 minutes and here it is.


Trains no longer rely on the driver alone – since the (SPAD induced) Ladbroke Grove disaster in 1999 further train control systems TWPS, ATP and ATP+ have been introduced.  These have reduced SPADs and reduced over speeding (another problem).  But they still happen, potentially serious ones at the rate of pretty much one per month.  Note that the UK still has the safest railways in Europe (excluding Ireland, who have a very small railway).

Train drivers are highly paid.  According to the average train driver salary is £47,000.  There are around 18,500 train drivers in UK (90% of whom are members of the ASLEF Union).  Therefore, the wage bill for train drivers is something like £870 million.  That is a big number in a capital intensive industry.

So the role of the driver is now to stop at stations and respond to signals before the TWPS kicks in.  Automating stopping at stations is not difficult and if that were done no drivers would be required.  Of course, the Docklands Light Railway is entirely driver less and operates to high standards of safety.  OK, its train speeds are not as high as mainline routes, but that is not the point.  Day in, day out the DLR works very well without drivers.

Driving a train may be well paid, but it’s also dull and, if some selfish muppet uses your train as a suicide device as they do at the rate of about one per day, potentially traumatic.  Trains hit about the same number of animals on the track too.

Why have drivers not been replaced?  One theory is that passengers like them being there.  This is baloney, as evidenced by the 110 million passengers that it carried in 2014-15.  Another is that the unions would not wear it – and we’ve all seen the disruption brought to London by rail strikes.  It may be that the technology of train control needs some further development = indeed there is a new EU wide standard coming in (I’m not biting this week).  But with the potential saving in labour alone I’m surprised that the government isn’t doing more.  Actually I’m not, as I can count on one hand the number of politicians that I have met who can read a P&L account.

Fully automated trains have the potential to increase capacity on the overloaded commuter routes through being able to run closer together, which might also reduce power consumption.  The technology exists and is largely installed and yet we are chasing the autonomous car.  Yes the technology exists and there is demand from the disabled (which is already being delivered).  But the integration cost is likely to be huge and it will require legislation.  Insurers need guidance too and I suspect that the software to enable an autonomous car to recognise an unautomated car being driven by an idiot, and when to worry about it, is far from simple.  Yes, cars have a shorter life and yes they are a consumer purchase and so attractive to the likes of Apple and Google. But they are a long way off and, last time I checked, we still have a budget deficit and provide £3Bn of support to Network Rail (net of the train operator’s payments of £800M).  Taking the driver out of the system would be a very significant saving.

But Sajid David got his headline and the Intelligent Mobility Find’s civil servants justified their existence.  The sad fact is that one does not need to travel to Brussels to find financial lunacy.