I work as a Project Planner for Network Rail in London.

I work free-lance, as many planners do, and have worked across a number of industries including oil & gas, civils, water and telecoms. But for the past 15 years I have worked in Rail. Although I have held a number of different project controls roles I tend towards project planning, and as I have signalling experience I am generally offered those positions due to the lack of experienced signalling resource on the market.

The reason for this intro is to illustrate that although I work in an engineering/construction environment, it comes with some unique aspects. CPA is pretty redundant, for example, as what drives Rail delivery tends to be access availability, rather than a critical path in the traditional sense. Likewise signalling delivery projects are a fusion of engineering and data production, which presents its own set of challenges.

However we are at the dawn of a new technology which will inevitably become standard, which is digital railway. Many countries have embarked on this technology and the UK is one of them. In broad terms we are talking about control systems which are managed by computerised systems to provide greater levels of management and the potential for more centralised and efficient control.

As a standard engineering/construction project we are used to a linear, cascading logic process ending in a go-live commissioning. This is represented rather elegantly by a Gantt chart, typically Primavera. But this process is not appropriate for data development which is far more cyclical than linear. Data is created, tested and given the results of the test, redesigned, retested &c. We can go from 70% complete one day, to 30% the next, to 90% the next. We are all aware how orthodox software developers do things, they release a version of software and wait for users to flush out the errors with help desk requests. But we cannot do that on the operational railway. If a software glitch turns all the signals red, or worse turns them green, we can't just press Ctrl+alt+delete. Trains smashing into each other is rather frowned upon in Network Rail.

So here is what I hope to discuss, how do we represent a hybrid approach to a digital railway project? Half software/IT and half engineering, half cyclical development and half cascading logic?

I really think that this type of project will become more typical as we see traditional control systems being increasingly managed by IT systems. From Signalling to SCADA. One similar example, although with key differences, is air traffic control. Does anyone have any experience in this or a related field?