Driverless Underground trains for London?

If TfL were building an entirely new railway, uncontaminated by past practice, then it would surely be completely automatic and more thinly manned than today’s lines. The technology is there to do it, and systems can be duplicated to provide very high levels of reliability. Passengers can be physically separated from the track with platform edge doors. Escape routes can be provided in larger tunnels with walkways. Good communications links will be in place, and trains could even be driven remotely. Staff and passenger culture can be created from new. oreover it is do-able, and can be seen to work elsewhere in the world.

Unfortunately, this is not the happy position in which TfL finds itself. It wants the benefits, but can only offer a system that is old and congested, not easily adaptable and has historic and cultural baggage that will take a huge battle of wills and serious money to alter.

FACT
We need to look at why TfL is currently entertaining the prospect of ‘no hands’ operation, but first we need to define some terms. The lines concerned are those deep Tubes soon due for upgrade, with what London Underground calls Automatic Train Operation (ATO): this means that all the driving is automatic, but there is an train operator on board who controls the doors from a driving cab at the leading end. We can extend this mode of operation, using technology to open the doors automatically and to use a roving operator to close them. I shall term this mode ‘Roving ATO’ (and this can be seen on the Docklands Light Railway). Fully hands-off operation is then possible where the train runs completely automatically, including door control (I’ll call this FACT to maintain association with the 1980 proposals). There are two flavours of this: FACT with somebody on board ‘just in case’ or ‘for reassurance’, and FACT where there are no staff on board at all. I shall term these FACT 1 and FACT 0 respectively.

Under FACT 0, railway operators will note that many complications go away. It provides unprecedented opportunities to introduce demand-led instantaneous scheduling that can be finely tuned to meet the various expected peaks and troughs in the service, and even demand that can appear at very short notice. The service controller presses a button and more trains come into service, just like that. The technology to do this already exists, and the benefits are significant, especially where disrupted train service can be corrected almost immediately. It also has the advantage of quicker turnaround times and minimal time wasted at stations, which can save trains.

These advantages are far harder to achieve on a staffed system because sitting behind every working timetable are duty schedules for all the staff involved (‘diagrams’ in main line speak). One cannot simply put unplanned trains into service: the extra staff have to be there. Their reliefs have to be planned ahead. Staff have definite hours of duty, and expectations of a break near the middle of the period. It is accepted that staff should be treated this way, but clearly it is a constraint to flexibility of operation. The cost of the train staff payroll is very high, as is all the scheduling and management resource that is required. Readers can see why a FACT 0 solution might appear attractive.

Moreover, looking back into history, we see that Underground drivers were once practically engineers as well. They had both to drive the trains and when something went wrong they were pretty much expected to fix it; of course they then had access to all the equipment. We have since seen equipment disappear from view so it cannot be got at. We have seen it become more reliable, so lack of familiarity with failure becomes a huge problem. We have seen trains become more complicated. We have seen automation, with the driver’s job now done by electronics whilst the person in the cab has become a traditional guard. From the perspective of those footing the bill, the inescapable conclusion is that they are paying more and more to get less and less.

Tolerance
With ever more intense and heavily used services, and ever more reliable equipment, when something does go wrong, are we beginning to ask too much of the operator? The system is becoming increasingly less tolerant of failure and it may be a good time to review the responsibilities of everyone who might get involved. It is a huge responsibility to find oneself alone on a train presiding over what could either be a minor, easily fixable occurrence, but where a missed symptom, unremembered procedure or just bad luck suddenly precipitates a major incident; there have been several recently which reinforce this suspicion. This is no reflection on the majority of staff, it is a systems problem that is also being faced in other industries, and perhaps the airline industry would suggest a comparison.

Even very reliable equipment is still not failure-free, and secondary failures can mask the symptoms of the first, with unfortunate results for the inexperienced. Traditionally, system operators devise additional processes and training to overcome apparent systems failure, which is expensive, inefficient and may not even work, especially when something really unusual happens. The outcome is ever more training for ever less frequent eventualities, pushing costs up, without necessarily the confidence that when something does go wrong all will be well. The consequences of a suboptimal response on a (frankly) overcrowded network could be very serious. Is an expensively-trained train operator in the driving cab the best solution to handling rare but inevitable failures? I don’t think anyone has an answer, but I do understand why the question is being asked.

Read the FREE sample issue of Modern Railways and the full Driverless Underground trains for London? article at pocketmags.com today: http://bit.ly/1DnltTo