Statistics show that signalling systems are becoming more reliable and this trend will continue as ongoing investment gradually transfers control to the new Railway Operating Centres (ROCs) being built in the UK.

That’s the good news, but with the increasing centralisation of control, failures can have a much greater impact than hitherto. Trains can be stopped for extended periods of time before permission is granted for them to move under temporary working arrangements.

With the railway now busier than ever before, the impact of stranded trains affects far more people than in the past and the build up of trains stopped at successive signals is a major concern for both Network Rail and the TOCS. Something has to be done to alleviate this situation.

The scale of the problem

Signalling failures can take many forms. Track circuit or train detection problems causing loss of train position information tend to dominate but power supply loss, points out of detection, level crossing malfunction and other failures all occur periodically. In recent times, vandalism and theft have been major causes with cable theft being a particular menace that accounted for 16% of all failures in 2011. With increased legislation to regulate the scrap industry and new initiatives from both the cable supply industry and the BT Police, this has now reduced to below 5%.

Delays of around two million minutes per year are caused by signalling failures. This has remained the statistic for some time for, whilst signalling technology has become more resilient, this is offset by the greater number of trains.

Arrangements within the Rule Book have been in place for many years to allow trains past a failed auto signal (i.e. a signal not protecting a set of points or level crossing) with the driver permitted to ‘drive on sight’ at low speed so as to be able to stop safely should any obstruction be observed.

However, a control signal that routes trains over a set of points or through a controlled level crossing cannot be passed until the route is proved to be in the correct position with points locked accordingly and a verbal instruction given to the driver by the signaller. This can mean getting someone to site to clip and scotch points and/or to have a person to manually control road/pedestrian movements at a level crossing.

Putting these arrangements in place can take a long time, typically up to two hours, and with the larger area of control provided by the ROCs, this time is likely to get longer. There has to be a way of streamlining the process.


The Combined Positioning Alternative Signalling System (COMPASS) seeks to take advantage of all the modern forms of communication that enable train position, direction and speed to be determined in real time for safety critical commands and information update requirements. These are then combined into a single data package that can reliably inform a signaller where trains are located should the primary signalling system have failed.

Sources of position information range from traditional track circuits and train detection through to GPS as needed for DAS (Driver Advisory Systems), train mounted camera devices used for track condition monitoring, RFIB (radio frequency identification), TMS (Traffic Management System) and maybe other technologies that will emerge over time. Not many of these are as yet standard fitments on trains and some will be slower to emerge than others. DAS is, however, being fitted to an increasing number of routes and trains and the first TMS systems will be in operation from 2016 at Cardiff and Romford ROCs.

COMPASS will not require every source of positional information to be available, but must be assured beyond reasonable doubt that a specific train is in the location that the data indicates.

This is fine but how does it help?

To minimise delay, the system must know the position of points and status of level crossings for an ongoing movement instruction to be given. This is achieved by the development of points monitoring units located at the lineside or in relay rooms to eavesdrop the position of points and report this back to the signaller using the public mobile 3G and 4G radio networks. Once this positional and status information is confirmed, a movement authority can be given for a train to proceed.

The time to validate the data is estimated at 10 minutes after which a movement authority can be given such that a train can move forward at speeds of 30 to 40mph. The authority may only extend to the next signal, also possibly at red, or it may be for a much longer block section. In the latter case, since train position information is constantly available, it is envisaged that COMPASS will permit a convoy of trains into the section under controlled conditions. In most cases where the extent of the failure affects only one signal section, the train will resume normal running if the next signal is at green. The time- savings are expected to be significant and should enable the railway to operate much more efficiently even though still in degraded mode.

The technology

An initial design, produced in house by the Network Rail Technical Services team, has now been manufactured as a pre-prototype by CHG Group in Sheffield. Essentially, the design embraces both a signaller’s unit, which will be installed alongside the normal work station screen, and a train borne unit to be installed in the driver’s cab. The signaller’s screen will normally be blank and will only be turned on when needed.

The point monitoring unit installed adjacent to the set of points has a SIM card to effectively make it a mobile phone. For the moment, this will just monitor the set of points and, every time they are moved, a call is initiated to the signaller to show the new position on his/her control screen. An integration unit will be at the heart of the COMPASS system that constantly collects all the train movement and direction data from whatever source is available and uses this data should a signalling failure occur.

manchester roc balcony(1) [online]

This early design will be deployed in the summer of 2015 on the Stoke Summit to Doncaster section of the East Coast main line (ECML), covering three sites at Ranskill, Retford and Carlton on Trent – all under the control of Doncaster power signal box. It is primarily to test the feasibility and integrity of the concept rather than overcoming any particular problem of signalling failures on that part of the line although, should such an incident occur, the practicality of the system will also be tested.

Wider system roll-out

In parallel to the ECML trial, Network Rail has carried out a number of feasibility studies as to how a wider deployment might be achieved and has issued an Invitation to Tender for industry to develop a commercial Compass Integration Unit and support elements. The specification for this is deliberately non-technical in order not to constrain tenderers from being innovative in their approach. It sets out the broad order requirements and the architecture that is foreseen. To date 13 bids have been received involving 58 companies including organisations from within academia.

A key requirement is the use of COTS (commercial off the shelf) products, not only to keep costs down but also to prevent bespoke developments locking Network Rail into a single supplier. From the 13 bids, it is anticipated that four will be shortlisted for feasibility studies and subsequently two companies will be invited to build and test a model system.

One of these will be chosen to implement a wider trial on a nominated section of railway. The trial area will be carefully chosen to test degraded mode operation in a location where signalling failures are known to cause major disruption to train service performance. From this the specification can be refined to allow a greater number of companies to produce products in due course.

The level of interoperability needed to ensure effective interworking between different pieces of equipment will be one of the factors to be examined. Once proven, it is anticipated that a wider roll out will be with a number of suppliers.

Longer-term deployment will need to be where benefits can be maximised, and this means a combination of dense traffic movements and complex junctions with many points and crossings. As such, it is envisaged that busy areas within the ROCs will be equipped but, in the medium term, many larger power boxes will also be provided with COMPASS to control critical sections of railway.

Future vision

It must be emphasised that COMPASS is not a substitute signalling system; it is there to enable degraded operation to be implemented in a much quicker timescale whilst maintaining the necessary level of safety for the movement of passenger and freight trains. However, with such a system in place, there could be many other applications where it would add value.

    • Whilst initially it will only monitor and prove that points and barriers are in a certain position, use of the communications link is envisaged to command a movement of these elements such that the intended route can be established as part of the degraded movement process.
    • Since COMPASS will provide a constant monitoring of all train movements in an area, supplying this information to lineside lookouts equipped with iPads or suchlike would be a great improvement compared to eyesight and horns/flags.
    • User worked crossings are an ever present safety concern and the provision of a local display device giving a ‘tactical picture’ of approaching trains and minimum safety distances would be a big advance compared to the present information provided by signallers that ‘a train is somewhere in the block section’.
    • Deployment of ERTMS will be totally dependent on GSM-R to communicate the movement authorities associated with ETCS. If the GSM-R system goes down, communication is lost and the trains will stop. Having a means of moving trains in this hopefully very rare situation will be important and COMPASS can provide this vital back up.

All these are for the future and one mustnot be carried away with too many ‘what if?’ possibilities. The adage of eating the elephant a leg at a time must be the way forward.

COMPASS is a 50:50 partnership between Network Rail and Future Railways (a sub set of RSSB) and the project has now become a part of the ‘Digital Railway’ concept. It has a budget of £3.9 million for the initial development and is expected to cost around a further £6 million for the operational trials installation and subsequent refinement. Points Monitoring has a safety case in place and has been put through the Safety Review Panel.

This is a classic example of ‘thinking outside the box’. New operating rules may be required but it must not be constrained by traditional signalling philosophy. Degraded mode operation has been a challenge for many years and over- zealous safety has made the problem more difficult. This is an opportunity to redress the balance for the benefit of all.

Thanks to John Collins, the project sponsor within Network Rail, for sparing the time to explain what COMPASS is all about.