Signalling technology and system engineering has moved on considerably over the past twenty odd years. However, it is still perceived by many as expensive and failing to keep up with technology advances that have emerged in other industries.

Network Rail is mindful of this criticism and has created a Signalling Innovations Group to investigate how new initiatives might be introduced at a tactical rather than strategic level. While the latter relates primarily to the ERTMS programme and traffic management and will take years to come about, more rapid improvements to existing engineering practices should be capable of being made that give quicker financial benefits and so promote additional investment. Ken Peters heads up the group, and The Rail Engineer met with him to learn how it operates and how it fits into the larger organisation.

Organisation and objectives

The origins of the Group date back to 2005, when two engineers (Ken Peters and Tim Franks) established a project known as STAMP – Signalling Tools and Methods Programme. At this time, new signalling projects were hampered by having to carry out planning work using traditional ways of working. Typical examples included the production of schemes plans, route and trackside surveying, signal sighting and the management of data for solid state interlockings.

This small beginning was given a budget of some £18 million to produce a series of modular tool sets, which could be used to not only get projects moving more quickly but to do so in a consistent manner. Good results prompted Network Rail to see if other initiatives might come about if the group were to be enlarged.

To be effective, the programme must be business driven. Every innovation must have a proven business case with the ones most likely to yield the best benefits being given priority. The newly formed organisation to deliver signalling projects, headed up by Mark Southwell (issue 107, September 2013) with Mark James as its professional head of signalling, was the logical place for tactical innovation activities to reside.

Signalling delivery, along with teleDsc_3924a [online]communications and energy supply, are treated differently to other Network Rail engineering activities which are organised on a Territory basis. There are several reasons for this – they are considered national business units, the Railway Operating Centres (ROCs) will not align automatically with territory or route boundaries and the need to maintain a critical resource mass.

Resources must be organised to produce maximum efficiency so that development work on a new project is not constrained by geographic location. This does not mean having everyone in one place and the Infrastructure Projects Group (Signalling) has offices in Birmingham, Reading, Croydon, York and Glasgow. However, keeping precious signal engineering expertise together as a single unit within the company was seen as essential.

It is also true that modern signalling embraces many other disciplines and thus the innovations team includes engineers and experts having specialist knowledge in mechanical design, electrification & power, civil structures, train equipment, data processing, business analysis and information technology, all being needed in modern signalling system design.

So how does innovation come about? It can be driven from a number of sources not necessarily within Network Rail. Getting ideas from the end user, be it the routes, the train companies or the supply industry, are the usual generators. Anyone who has a ‘what if’ question will be welcome to present the group with a challenge.

After assessment for viability and business gain, the solution may progress either from within the group’s own expertise, or in conjunction with the supply industry, or in collaboration with universities. Examples of the latter are ‘The Digital Railway’ project with Oxford University and the ‘Guidance, Navigation and Rail Vision’ project that emanated from Loughborough University.

Driven by efficiency gain, it sometimes means higher cost in the short term to obtain improved output in times to come.

Achievements so far

With STAMP being expanded to cover other things, ongoing work has a new banner – SNIP (Signalling National Innovations Portfolio). Remembering the remit to reduce the cost of signalling, the use of COTS (Commercial off the Shelf) equipment features in much of the work being carried out.

Plug and Play Cabling (issue107, September 2013) originated from the Infrastructure Projects Group (Signalling). However it was not just the design and production of plug ended cables, much more a whole re-think on how trackside installation of signalling equipment
is achieved. Measurement, design, procurement, on site storage and testing methodologies all have to change, and some of these still require attention.

Attempts to speed up cable route measurement by aerial survey and train based survey did not yield the desired accuracy. Recourse to a person with a measuring wheel has been necessary, somewhat ironic given that the whole concept is to reduce the number of man hours needed at the lineside. More work is therefore being done to see if a combination of measuring techniques can be used to achieve the desired end result.

The adoption of Class II trackside power supplies has similarly had implications well beyond just replacing a three core cable with only two cores. Double-insulated systems need new designs of transformers, terminations and end usage equipment, all developed in conjunction with the supply industry. Just in case anyone is worried that earthing has been abandoned, lineside location cases, equipment racks and all other metal work is still safely bonded to ground potentials.

Exploiting data

Asset capture and recording has long been a problem for the rail industry and criticism from the ORR that Network Rail did not know what it owned has produced a number of initiatives to get this corrected. The SNIP project ‘Data Exploitation’ is building on the previous upgrade of the Track Recording Unit, which has seen train-borne cameras upgraded and re-positioned to meet the need of signalling asset surveys. The semi-automated process is programmed to identify any asset needed for a signalling scheme ranging from signals and point machines to block joints and axle counters.

The raw data currently has to have a human interpretation (known as profiling) that can then be used to develop a signalling plan. It is hoped that an automatic image recognition system can be devised to minimise human intervention. This will be achieved by building on a partnership with BAE Systems at Yeovil, so adding weight to ensure the concept can be successful.

Omnicom's OmniSurveyor3D cameras on the Network Rail Track Recording Unit.

Omnicom’s OmniSurveyor3D cameras on the Network Rail Track Recording Unit.

To date, 30% of the railway has been surveyed for signalling equipment. With the concept developed, the ongoing work will be taken over by the ORBIS (Offering Rail Better Information Services) team who are charged with keeping the asset database current.

Data preparation has been a problem area ever since computer-based signalling was introduced. Getting some form of standardisation is the goal and a good starting point has been the automating of scheme plan data. Exporting the proposed layout in a standard data package (Signalling Data Exchange Format) using XML mark up language has been achieved and this in turn will enable the information to be exported to suppliers to speed up the production of data for future interlockings.

Another data validation exercise has been the verification of signalling rules in the new modular signalling projects and a trial involving Siemens (ex Invensys) using a product called iProver shows promise. Proving that SSI (Solid State Interlockings) data is safe has been a challenge from day one and even today, it is not entirely a precise science. The innovations team have developed a tool called SSI Workbench that is a networked version of the SSI work station for preparing such data. It is not the whole answer to the problem but is one of many ways in which the safety approval of data can be streamlined.

Future Work

There is a shopping list of work for the Innovations Group to do under the SNIP banner and the list will continually need updating as initiatives arise.

Modular Signalling is likely to be adopted for the signalling of secondary routes. This technology has the potential for much reduced cost but can the planning work be made easier? A tool called ISP-Sketch is capable of producing the associated simplified scheme plan and then exporting this for data preparation. It was not used on the two trial installations but is likely to feature on future projects.

Signal Sighting is already automated to some extent. Work is underway to carry out image modelling using both video footage and/or virtual reality so as to achieve desk top sighting without anyone having to go trackside. The benefits will include consideration of options so as to get the preferred end result and then to use the model for assessing signal construction requirements and driver route learning.

Level Crossings are an ever-more-expensive item in railway infrastructure, so how can their costs be reduced? The use of industry standard PLC equipment (programmable logic controller) for the control circuitry is one such suggestion and is being progressed in conjunction with the Network Rail central signalling team.

Design Tools which take the tedious workload out of signalling design is a worthwhile objective. In conjunction with the signalling industry, the development of automated processes to produce scheme plans, headway analysis, signal spacing and data preparation has funding approval for the upgrade of existing tools.

Product Acceptance is a complex process within Network Rail and it can take months to get even the most basic of products approved. Is it all necessary? Many think not and a more pragmatic approach is called for. The identification of products that can make a real contribution to efficiency must be prioritised.

Determining products that sensibly do not need formal approval is another challenge. Cataloguing items by complexity such that anything that could be bought as a standard piece of hardware would not require formal approval will be a sensible step forward. Equally, if a product is successfully in use elsewhere in Europe or the wider world, does it really need to go through the complete approval cycle? Getting progress on this latter issue might help prevent the bespokery that so often takes place.

So, there is nothing earth shattering in what this group sets out to achieve. It is just the application of good plain common sense and logical evolution by engineers who have been around in the rail signalling business long enough to know where the scope for improvement might exist. They do not seek to work in a vacuum and interchange of ideas with other bodies is very much part and parcel of the process. Long may they continue to flourish.