No, not a new way of finding where you are going, but a review of the Network Rail and RSSB joint project to find a way of getting the railway operational again, should a significant signalling failure occur.
With Network Rail’s infrastructure being required to provide ever-more capacity in terms of number of trains run, the impact of signalling failures can be very significant. Trains queue up, one behind the other, whilst someone is sent out to site to rectify the problem or arrange for hand signallers to be put in place, maybe also securing the points.
Mitigating this problem formed the main element of the initial COMPASS project, which has now been termed the Degraded Mode Working System (DMWS). COMPASS DMWS is a project aimed at reducing the time to get trains moving again and was first reported in issue 113 (July 2015). Since then, a greater degree of realism has entered the thinking and a meeting with Chris Fulford, the project’s lead engineer, revealed the present progress.
In simple terms, the project will design and develop a system whereby, in the event of a signalling problem, an instruction can be given to a train driver that it is safe to proceed beyond the failure locality to a distant position determined by the signaller. In effect, this is an electronic version of the setting up of present-day Temporary Block Working (TBW), with its associated paper-based instruction giving authority to pass signals at danger. The project will be applicable to both conventionally signalled railways and lines equipped with ERTMS/ETCS. It is not currently part of the Digital Railway initiative, since the technology solution is at an early stage of development.
At this stage, the failure situations are likely to include loss of signalling power supplies, loss of indication at the signalling centre, multiple track circuit/axle counter failures, lineside cable damage and theft. DMWS will initiate checks to ascertain that nothing untoward exists that would prevent a train from proceeding safely.
To achieve this, DMWS must know the accurate position of the train, which track it is on and the intended route it is to take, plus the whereabouts of other trains in the vicinity. DMWS must also know the status of critical infrastructure, such as the lie of the points and whether they are locked in position, or whether the level crossing booms are up or down and road lights are working.
Once these facts are proven, then the signaller can give the driver an instruction (this could be sent within an encrypted SMS text message acting as a data packet) to proceed to the specified distant position. This will be known as an ‘Authority to Move’ (ATM), so as to avoid confusion with a Movement Authority (MA) as used in ERTMS/ETCS train control.
The means of locally and independently assuring the precise state of a set of points or a level crossing will be achieved by having a COTS (commercial off the shelf) PLC (programmable logic controller) linked to sensors that indicate the points or level crossing status. The PLC will connect to the central system using the GSM-R network radio, fitted with an appropriate SIM card, linked to modems located in the DMWS ground equipment. The radio will be interrogated by the DMWS central system with a separate control screen being provided at the signaller’s work station, which will normally be switched off and only activated when required.
Initially, the PLC will be programmed only to inform on the current position of the ground equipment, but more adventurous commands are foreseen once the system is proven. The PLC will constantly receive points and level crossing status information, not just when a failure condition occurs. It is possible that the GSM-R system might also have failed, for instance if a cable has been cut that feeds the nearest base station, and, to guard against this, the radio unit will comprise a public mobile receiver as a secondary unit, with the combined unit being a dual-band receiver.
So much for the trackside and control centre units, but what will be needed for the train-borne kit?
A means of communicating the ATM instruction to the driver has to be available. This can either be a free-standing unit in the cab or, more likely, the COMPASS DMWS facility could be built into existing systems such as the train borne DAS (Driver Advisory System) unit or perhaps the next generation GSM-R voice radio. The latter will have considerable computing power and a Rail Engineer article in issue 129 (March 2014) described the additional uses that might exist in the new generation of train mobiles, including facilitating DAS information on its screen.
DMWS will need train position to be reported to the signaller on a near-constant basis and this implies the use of GNSS (Global Navigation Satellite System). The new generation of train radio will have a GPS input, and thus the feed for this will be readily available. It is, however, the intention that the train and freight operators will be free to choose whatever equipment suits their own fleets.
Whilst DMWS is seen as a natural addition to the operational management systems within the new Railway Operating Centres (ROC), there is no reason why it cannot be deployed in any IECC (Integrated Electronic Control Centre) or power signalbox. Indeed, if the expected advantages are to be realised, the pressure will be to deploy it as widely as possible.
Progress to date
The 2015 intention to pursue a trial on the ECML did not materialise, beyond fitting some test infrastructure monitoring equipment to the three nominated locations between Doncaster and Peterborough.
A more pragmatic way forward is now progressing, with a three-part R&D programme. Part 1 consisted of an invitation to 13 companies, issued in early 2015, to bid for producing a feasibility study on how a DMWS system could be progressed.
Five of the 13 companies were awarded a contract in September 2015 and invited to submit a further bid for Part 2, entailing production of a laboratory-based simulator to demonstrate how a DMWS would operate in practice, including the integration of the various sub-systems.
From the feasibility studies and proposals received in Part 1, and after the relative merits of the responses had been assessed, two companies – Thales and Altran proceeded to build a simulator, Altran in Bath, Thales in Cheadle Hulme, the work taking place between March and December 2016.
Additional studies were included in the work packages, such as RAM (Reliability, Availability, Maintainability), safety and security, and human factors implications. Both companies were also required to submit a proposal for Part 3 – the building of a functional DMWS demonstrator.
The two simulations, together with their associated deliverables and the proposals for Part 3 were duly assessed, with Altran being awarded a contract extension for Part 3 in February 2017. This will involve provision of a real demonstration on the Hertford Loop test track. The line is double track, but the infrequent train service allows one of the lines to be used outside of peak hours for testing purposes and it is where the ETCS integration / interoperability tests have taken place. The single line has no points or level crossings but these will be artificially inserted into the test section using the real points and level crossings that exist in the Network Rail Walsall training centre – it’s amazing what can be done with Network Rail’s in house broadband transmission links!
The test train used for the ETCS work (a Class 313 EMU) will be fitted with the train-borne DMWS equipment, including the necessary GPS satellite tracking picture and ‘distance to go’ count down. The signaller’s control panel will be fitted in one of the test control rooms for the line, located at the Hitchin ERTMS National Integration Facility (ENIF).
GPS co-ordinates for the relevant track geography (such as signal locations) will be provided by translating video images that already exist of the railway using the Systems Data Exchange Format (SDEF), thus enabling the infrastructure information to be linked to the GPS location.
A two-week trial slot is allocated for January 2018. During the test period, the train driver will not be expected to directly take part in the trial as this may cause confusion with other test programmes. The DMWS ATM instructions will be monitored by Altran test engineers, who will instruct the driver accordingly.
Altran will have the responsibility for proving that the combination of GPS positioning, points-lie information and historic berth occupancy is capable of determining which track the train is on. It is hoped that the on board requirements can be largely incorporated into the new Siemens GSM-R voice radio which will have 4G and WiFi capability as well as GPS positioning and a modem, thus being able to switch from GSM-R to other radio modes.
Since the Hertford Loop site already has a safety case as a test track, the demonstration work can be carried out without any new safety requirements needing to be specified. No trackside infrastructure needs to be fitted, as this will exist in the safe environment of Walsall training school.
DMWS will need to effectively disconnect the point machine or barrier mechanism from the interlocking, as to initiate commands from two sources during the period of disruption could be confusing and compromise safety. DMWS may also suppress any TPWS operation within the failed area so that trains will not have to stop for the driver to manually isolate the on-board TPWS equipment.
Following this, an operational trial should take place on a chosen piece of railway, which will then need a separate safety case to be produced. However, since DMWS is essentially an electronic version of the long-established manual means of setting up temporary block working, it is anticipated that the safety procedures used for the latter can be the basis of the DMWS intentions.
There can be little doubt that, if successful, COMPASS DMWS will become an important tool to get trains moving again more quickly when degraded operation is necessary. With the ability to assign the length of the DMWS block section, and with the train’s movement regularly updated, it will give the signaller much greater control than the manual ‘man on site’ situation existing currently.
It must be emphasised that the system is only an aid to operation, recognising that the signaller and driver remain in control. DMWS is essentially a way of enhancing the communication protocol in getting an instruction to move a train from point A to point B.
Once the system is proven in operation, the ongoing vision extends to using DMWS to initiate an instruction that would move a set of points for the intended route and for level crossing activation to commence. Indeed, the level crossing sequence at Walsall has already been activated by a similar type of remote command, as have points at Doncaster training school. However, that is for the future, and it is very much one step at a time for now.
The immediate ongoing work will be to complete the practical demonstration, produce a business case and the writing of a procurement specification for an operational trial. All these will need to embrace the use of COTS equipment as the justification for proceeding will need to minimise the amount of financial outlay required.
The total number of DMWS trackside units to be procured could be considerable as one unit is only expected to monitor one or two sets of points or one level-crossing site. An analysis carried out at Didcot showed that up to 36 units would be needed to cover all the various junctions.
DMWS deployment will need to link into the railway’s business performance model and will be prioritised to the most vulnerable inter-city and major city suburban routes.
Rail Engineer will continue to monitor the development of this fascinating project and will report progress from time to time.
This article was written by Clive Kessell.