Why technology usually over-promises and under-delivers

We have had railways moving people around for over 150 years in the UK. From humble beginnings, we now have over 5000 miles of track serving to within perhaps 50 miles of virtually everywhere in Great Britain.

In these early days, railways were a difficult and inefficient system, mainly restricted by the available technology. For instance, steam trains could not travel very far without re-filling the water, something that was available at most stations and could take a noticeable amount of time, some larger engines were equipped with scoops to collect water from track-borne troughs but these had their own maintenance costs, especially if someone broke them by failing to retract the scoop!

Coaling was an issue and although diesel locomotives also require fueling, coaling was extremely variable from the largest semi-automatic filling machines at larger depots to literally a coal bunker and a spade at others, this was another potentially very time-consuming exercise.

Most larger steam locomotives usde a tender to carry coal and water and had to be turned so they were running locomotive first. Although the turning itself was not massively slow, in almost all cases, a loco would have to be released from a platform, driven to somewhere that was not necessarily very close, amid all the other services, and then turned before it could be brought back and used. In some cases, this could be performed around a triangle junction but this could take 20 minutes, even if there were no delays or stops.

Starting a locomotive from cold was a very slow process, taking anything up to 8 hours from cold to having enough steam for service. Naturally, this could not be started until all maintenance was completed which is why we used to see long lines of locos on depot departure roads, just waiting to raise steam.

Let us also consider the signalling system. All signals were originally mechanical and had to be maintained/lubricated and almost all signals as well as points were mechanically linked to a signal box. All of these linkages also had to be manually inspected and maintained as well as repaired. For purely mechanical reasons, these boxes would generally have to be within about 1000 yards of the points or signals they were controlling meaning that boxes were placed at very quiet locations, that happened to have a level crossing or a siding and these would require at least 1 person per day to staff. Many other boxes were staffed 24 hours per day, sometimes by as many as 6 people. The total number of signalmen in the 50s must have been massive.

So, as with most systems, the railways have improved, thanks to "technology". In the UK most trains are now electric, none of them need turning, very few still use locomotives anyway. Watering is only for toilets. Fueling is handled by pumps and hoses and is much easier and cleaner than coaling. Most trains spend their working day literally running end to end with no breaks for maintenance, turning, coaling, fire-lighting or anything else. An electric train can be started from cold and pulling a train within about 10 minutes. These improvements also mean we don't need a driver and a fireman, we can just have a driver, that has halved the salary budget already! Maintenance is far easier on electric trains than steam locomotives, most of which was purely mechanical and required welding and boiler-smith repairs.

Then we can look at signalling. Virtually all signals are now electrical and almost all mechanical signals that remain are electrically operated remotely. These are all controlled from power boxes requiring a tiny fraction of the people that used to be employed - perhaps 3 people covering 100 miles of railway.

And yet, the railways are not so much cheaper that tickets are half the price they used to be! How is this possible? How can we be saving what must be over 50% of the wage bill and enormous amounts on maintenance and dead time while we move locomotives around? How can we require far fewer locomotives since we don't lost time waiting for steam to be raised and coaling and watering to be carried out but still trains are considered a luxury by many people.

It's because technology over-promises and under-delivers. What actually happens is that when we are sold new technology, we are told that it can save us a packet, so it is charged out at a packet. Mechanical signals are a pain, but once they are installed, it is just maintenance costs - a bit of grease, some bits of metal and bolts to repair. What about a colour-light signal? I imagine they probably cost a few thousand each, more for the elaborate ones and then the installation is horrifically expensive. Specialist railway contractors are required due to Health and Safety jobs-worths who use the railway as a giant cash cow (either maliciously or perhaps well-meaning but naive). These contractors then have to pay money to get a line possession and perhaps some other specialist equipment like track trolleys - obviously people couldn't just carry things from their vans any more. Once they are installed, they work fine for a while, but what happens when they go wrong? No bits of metal and bolts here! Just another call to some specialist contractor who has to have specialist knowledge in safe systems of work, probably costing you £1000 per day (when they get paid £25K per year!) just to be able to go and buzz some wires and work out that a fuse blew or a bulb went out.

Look at all these other areas and there is something, some new technology that sells itself too well and which people buy into but without the knowledge to know whether the investment will ever repay itself. I used to work on locomotives that had a fairly basic microprocessor which logged faults. Guess what? Training was expensive, changes were basically out of the question and for most problems, we had to go straight to manufacturer to get help which was either slow or expensive or both.

Very sad really. I think many people still do not understand how technology is supposed to make things better/easier/cheaper and not provide license for someone to make a packet off the back of a well meaning company.