There was a Railtrack initiative to broaden the signalling supplier base, which led to Siemens setting up new teams to customise their German Simis W equipment for British use.
Railtrack HQ decreed that SSI, the interlocking system that had been jointly developed by BR and it's two main UK suppliers, was a dead duck. They were not going to buy any more, it was too expensive. In future, they were only going to buy cheaper, off-the-peg overseas systems.
The big problem was that was not what their engineers wanted. Those on the Southern in particular were not happy at having two non-standard trial systems foisted on them. The result was that the suppliers were caught between a rock and a hard place. While Railtrack HQ wanted off the peg systems, their engineers were demanding that they be reverse-engineered to behave as much like SSI as possible. It was like buying a VW Polo, and reverse engineering it into a Rolls Royce - you will spend far more money and considerably more time than if you had just bought a Rolls Royce in the first place. And that is more or less what happened with these trial schemes, they went way over budget and over time while the systems were extensively re-engineered.
It also didn't help that Railtrack wanted the UK data structures developed from first principles, rather than copy from existing relay or SSI practice. They wanted a clean break from the bad old days of BR.
And when the schemes went way over budget, and didn't seem to be giving the savings hoped for, then Railtrack HQ lost interest. And with nobody in Railtrack pushing them, the schemes withered on the vine. The two systems that had been commissioned were got into a "usable" state, while the third (ebilock) was abandoned before it could be commissioned. There was no incentive to do any more.
I was working for Adtranz at the time, who were doing the ebilock pilot. I had worked on various overseas ebilock schemes, so knew the system. Ebilock used an architecture very similar to SSI, with distributed lineside object controllers. However, being more modern, the object controllers were a lot more sophisticated than their SSI equivalents. Both hardware and data design were a lot simpler and quicker, as the design tools were more sophisticated, using mainly templated designs and data. Construction too was quicker, as it used plug and play wiring assemblies, with almost no free wiring. The system showed a lot of potential. Our office in Stockholm could turn out ebilock schemes far faster than we could using UK technologies of the day.
However, the adaptation had gone way over budget and time, for the reasons already described. And their Southern Zone engineers didn't want another "Portsmouth" on their hands. So the plug was pulled.
The amount of manhours and money that had been expended for nothing doesn't bear thinking about.
www.networkrail.co.uk
