Archive Section

The Locomotive & Carriage Institution
Home Centenary 2011 Continental Visits Lectures Newsletters History Visits Pictorial Visits Narrative Seminars
Signalling & Communications in the 21st Century:

By Anthony C. Howker

Immediate Past President Institution of Railway Signal Engineers

Technical Manager - Train Control Systems, West Coast Main Line Route Modernisation Project, Railtrack plc, United Kingdom

(This article first appeared in the June 1996 Newsletter; obviously signalling and telecommunication technology has continued to evolve since then.  As you maybe aware, the plans for the signalling system for the West Coast Main Line [WCML] route modernisation have changed since this article was written and, for the time being, more conventional line-side fixed-block signalling is being utilised.  The new GSM-R network referred to under WCML is, however, being rolled out by Railtrack on the WCML and then nationally over the next few years.  This system will replace all existing radio systems such as CSR, NRN, ORN and ETRB.  GSM-R will have greatly improved functionality compared with existing radio systems and will be able to detect train position with out necessarily using track circuits.  Tom Chaffin, Telecoms Engineer, Railtrack, Thameslink 2000 Project).

History:

The development of Signalling Systems on Railways had been evolved with the need to provide interlocking between points and signals, and block working to keep trains a safe distance apart.  (Trains also needed automatic Brake Systems on the whole train but for the purpose of this paper are assumed to exist!)  Systems for the control of single lines also needed to fulfil the above requirements, but the effects of a collision were felt to be especially severe and therefore required extra protection.  In principle it was developed using Semaphore Arms as the medium of transmitting information to drivers, worked mechanically from interlocked mechanical lever frames and used an electrical telegraph block system to carry information along the line as to the clear or occupied state of the absolute block section.  The Semaphore Arms were arranged to give information to the driver as to the state of the line ahead and the route on which the driver had permission to proceed.

Of course, it was always assumed that drivers knew all the intimate details of the line covering all speed restrictions, positions of signals and the maximum speed of all turn outs on all routes.  In principle, this signalling practice did not alter in any fundamental way from about 1880 until the late 1960’s.  Of course, various upgrades were installed such as the track circuit and block systems, which were connected into the system electrically supplementing the mechanical interlocking.  Colour light signals were also introduced from the 1920’s, but in principle following either the original “Route Signalling Systems” or “Speed Signalling” - a system that told a driver the maximum speed he could travel.

The controlling medium gradually changed from mechanical to electromechanical interlocking, then to panel control via relay interlocking and at the present time solid state or computer interlocking: managed by Integrated Electronic Control Centres (IECC) with Automatic Route Setting (ARS) by computer.  The information given to the driver in “British” type signalling continued to be on the basis of telling him where he was going and how many block sections were clear in front of him.  This Route Signalling information was conveyed to him by colour light signal heads supplemented by junction indicators in some instance giving alpha/numeric information.

The early railway lines built overseas were as a general rule built by British engineering companies, and the first signalling used was Route Signalling, as in Britain.  However, in those countries where the British influence waned speed signalling became the norm, especially in Europe and North America.  Their argument followed familiar lines, Speed Signalling did not need drivers to have route knowledge in intimate detail, although it was recognised that some knowledge was essential.  Over the years conventional interlocking and Absolute Block working became mandatory, as in Britain, and only in the aspects given to drivers was there a difference.  In countries where the British influence remained strong route signalling with, some minor details differences, was perpetuated.  With the advent of power signalling, many countries who previously had installed route signalling, took the opportunity of introducing Speed Signalling, most of them following North American standards which were set in the early 1900’s.  Some examples being Argentina, South Australia. Victoria and New Zealand.

The principle of Speed Signalling is to give a driver advance notice of the speed he should pass the next signal.  Usually, but not always, the speed related to junctions, but as the system grew, it was also applied to fixed speed restrictions.  The problem realised early on was the shortage of signal (aspect) indications to give enough varieties of speed.  Some authorities in Europe eventually had to use characters to give enough speed variety, but in North America, the decision was taken to standardise on a few speed settings, and thus keep the signal aspects to a minimum.

Methodologies Utilised:

Interlockings:

So far we have briefly touched upon the History of Signalling as applied (or shown) to the train driver, but of course, engineering methods varied from country to country ranging from Relay lnterlockings to Solid State.

Even relays differ, most European countries   use metal-to-metal contacts whilst North America uses silver-to-carbon contacts, as does France and Britain.  Other countries around the world tend to follow their original supplying country for historical reasons.  Therefore India, Australia, Malaysia. New Zealand and other Commonwealth countries followed British practice in their interlocking solutions, even if they now use Speed Signalling rather than Route Signalling.  Modern installations have now progressed to Solid State lnterlockings with British SSI in use now in Denmark, Ireland, Spain, Portugal, Central and South Africa, Hong Kong, India, China and Australia.  These are unique in that SSI was a joint development by British Rail and two of their major suppliers.  Elsewhere Solid State lnterlockings are in use, but totally developed by the supply industry, although no doubt their home’ railways had a say in the development.

Train Detection:

Train Detection throughout the World continues to be Track Circuit based, although the form of Track Circuit varies tremendously.  Most countries that base their signalling practices around AAR specifications use a track shunt value of 0.6 ohms which allows them to have very long Track Circuit’s in non-electrified territory.  Much use is made of Coded Tracks, i.e. a DC Track Circuit turned on and off at varying rates.  The most usual rates being 80, 120 and 180 cycles per minute.

These codes not only give a good track circuit length, but also can be used for other purposes, such as aspect sequence and approach locking. Receivers mounted on a loco can also give Cab Signalling and Train Protection.

Originally the coding was applied (and still is) by Pendulum Mechanical Transmitters, but over the past 20 years, electronic means have been used to transmit code.  The modern versions of this form of Track Circuit now use computers to apply digital signals to the track circuit, and allow a far greater number of codes to be used.  Countries with coded Track Circuit’s include North America. Holland, Italy, China, Australia and Spain.

Other forms of track circuit in general use are obviously the 50 or 60 Hz type with Impedance Bonds for use on DC traction.  This type is common in Europe, Northern America, India, Asia, Australia and China.  The GASL high voltage impulse and Westinghouse FSK track circuits are also used. These are all jointed track circuits.

When it comes to AC traction, DC track circuits which are AC-immune track are common, but in areas of 25-0-25 KV traction, the jointiess track circuit is becoming the norm.  Many variants exist, the well known ones being manufactured by GASL, CSEE, WB&S, Siemens, Ansaldo, US&S, GRS and Safetran.  They all use the principle of a tuned circuit at the extremities, and some are capable of length extension by the use of capacitors across the rails at regular intervals.  Many of these track circuits are capable of being modulated and thus utilised for cab-signalling and ATP/ATO Systems.

In countries such as North America, South Africa, Thailand, China and Australia, railways have been constructed over vast distances, are mostly single line, and passing loops are great distances apart.  In these instances, the block sections between loops are often fitted with Axle Counters.  The principle of an Axle Counting System is very simple - count the axles on a train that enters the Block Section, and then count them out again as the train leaves.  As long as the resultant count is zero, the section is clear.  Original Axle Counters used mechanical treadle depressed by the wheel flange, but all modern version now use electronic means to detect wheels.  With the advent of modern communication systems and safety critical software, today’s axle counting system is not limited to just counting axles, it also acts as a fail safe transmission system between passing loops, and the medium of transmission now uses fibre-optic, PCM systems, radio and satellites as the bearer for the information.

Of course the final Train Detection System in use for long single lines with low-density, but maybe heavy traffic in weight is VOICE!  Modern Train Order Systems are in use in North America, South Africa, Australia utilising radio and computers at a control centre.  Sometimes the order to proceed is hand written, or printed out in either the loco or a local control point.  They all rely on voice communication from the driver to confirm the train is clear of a section, but modern techniques are now being used such as Train Complete Systems and transponders for positioning.

About this time the first use of VDU’s for indication and control appeared in France, Germany and Italy, and quickly were utilised in other parts of the World.

Nowadays Control Centres without computers are simply not installed.  Modern systems such as IECC (Integrated Electronic Control Centre) not only display everything the controller wishes to see, but also acts as a disseminator of management information to all parts of the railway, such as passenger information, train crew details, train consist etc.  The IECC also routes trains automatically using timetable information and a control strategy for solving conflicts.  By these means control of large areas can be accomplished by minimum staffing.  In other parts of the world such as North America and Australia, complete Railroads are automatically controlled from a Central Point using Electronic display technology not only to show actual position and routing of trains but also such things as time/distance graphs in real time which are capable of projecting forward in time to show conflicts and their possible resolution.

Control Centres:

Control Centre technology is probably as varied as the number of control centres in use!

The first control centres were signal boxes, “Interlocking Towers” where the control exercised was local to the geographical position.  The controlling medium of the Signalling System was originally mechanical levers for switches (points) and signals.

Early changes to the operational medium consisted of miniaturising the levers, and eventually changing from mechanical to electromechanical interlocking.  Finally, the levers turned to switches with all electric interlocking.   A the same time train detection (or positioning) went from direct visual (out of the window!) to the use of Indicating Track Circuits occupied on an illuminated diagram.  The operational medium remained “Unit Lever” in many countries, but as the areas of control extended some (but not all) countries changed to a Route Setting” control medium.  Europe, including Great Britain, quickly moved in this direction and often indicated the Route Set function by means of white lights on the diagram from the controlling signal to the exit signal.  Especially in North America, areas of control became vast (long single lines) and in the 1930’s Control Centres became established to control Centralised Traffic Control (CTC) systems.  The relay interlocking were dispersed along the railway lines at station crossing loops and were controlled and indicated using a pair of wires, usually carried on a pole route and a relay coding system.  Even today, such transmission systems are known as “Coding Systems”.  As railways throughout the World rationalised and integrated their control system, hardwired control and indication panels became a display system of the past.   Early advances in computer technology allowed more “management” functions to be commissioned as distinct from “signal” functions, such as Train Identification, and Automatic Route Setting by timetable etc.

Great Britain was the first country to use electronic TDM (Time Division Multiplex) Systems in 1959 for the remote control of interlocking and Europe quickly followed suit.

CTC systems in Northern America, South Africa and Australia also recognised the benefits of Electronic TDM Systems, and the first computer variants appeared in 1978.

Communications:

Throughout the history of Signalling (probably better described nowadays as Train Control Systems), corresponding developments of communication systems were also taking place.

The industry has benefited from the explosion in the commercial world of communications from telegraph instruments, teleprinters, 3 channel carriers onto microwave systems, electronic exchanges, PCM Transmission Systems, fibre optic cables, cellular radio, satellite systems and mass information links.  These systems have been introduced in the day to day operations of our railways with little disruption and often wonderment on how we ever managed beforehand!

Automatic Train Protection:

During the history of Train Control, it was quickly realised by engineers that as a system for giving information to the drivers of trains, we relied utterly for safety of the system on the knowledge and discipline of those drivers.  We needed to complete the control circle by enforcing the authority to proceed by some means.  This in turn led to the development of Automatic Train Protection (ATP) Systems.  Early systems used mechanical connection between train and signalling, moving onto coded track circuits and cab signalling.  Metros around the world were the first railways to become deeply involved in such systems, and have led the way with Automatic Train Operation.  Modern systems (e.g. Queensland Railways) now use radio and passive beacons for ATP implementation and are using the techniques that are used in transmission based Train Control Systems.

The Future:

Train Control Systems:

As has been seen, Train Control Systems have moved in step with the development of technology both in Signalling and Telecommunications.  The advent of these new technologies has given the railway industry the opportunity to take a quantum step in the signalling and control of trains.  The United States of America were the first to use such systems, driven by the economics of signalling long stretches of their railways through sparsely populated countryside.  Advanced Train Control Systems (ATCS) were specified jointly through the Association of American Railroads (AAR) and Industry in the 1980’s, and laid down methodologies using combinations of radio and train positioning systems as the basic requirements for Train Control.

The major benefits to the Railroads came from extensive radio coverage and data transmission between locomotive and Control Centre covering such diverse requirements as “authority to proceed” and locomotive condition monitoring.  It also helped change operational and maintenance procedures.  Similar systems in principle have been tested in Canada, Africa and Australia.

The General European View:

Within the European Community, the executive arm is the European Commission (EC), and since 1990 the EC has made great efforts to move towards technical harmonisation of Train Control Systems in order to facilitate a Trans European Network (TEN).  A lack of common ATP standards and differing signalling systems in various countries has inspired a European Train Control System (ETCS) project.  The roles of both the Railways and Industry in the project have been defined with the European railways being responsible for the definition and control of the Functional Requirements Specification (FRS) and the System Requirements Specifications (SRS) of ETCS.  The signalling industry has formed a nine company consortium called Eurosig which is responsible for elaborating the interface specifications for all the systems to ensure that different manufacturers’ equipment is compatible.  These are the so called Form Fit Functional Specifications (FFFS).   All the specifications when produced will become public domain documents for use in tendering.

The British View:

Whilst British Railway Industry and Railtrack are part of the EC Standards setting group, it was recognised that the timescales involved did not suit the problem facing Railtrack in its need to resignal, refurbish and upgrade a very important part of the British Railway Network.

West Coast Main Line:

The West Coast Main Line (WCML) is Great Britain’s busiest mixed-traffic railway corridor.  It runs from London Euston to Birmingham, Manchester, Liverpool and Glasgow and also connects with Edinburgh.  The line crosses 16 counties or regions with a population of some 16 million.  It connects to key towns and cities in the West Midlands, the North West, North Wales and Western Scotland as well as to the ports of Holyhead, Liverpool, Heysham and Stranraer for Anglo Irish traffic.  The line is used by more than 2,000 trains per day, carrying both passengers and freight.   Annual traffic is approximately 5 billion passenger-KM and 5.5 billion freight gross tonne-KM.

Originally built in the 1830’s and 1840’s, the WCML was the first inter-city railway in the world.  Parts of the line were widened, to four tracks, in the first half of the century, and the line was electrified between 1955 and 1975.

Following a feasibility study, plans have been put in place for the modernisation of the WCML including improvements to track infrastructure, replacement of the existing colour light signalling by a new Train Control System (TCS) and implementation of a new Management Centre.

The Train Control System will replace the existing control system and will ensure the safe passage of trains according to Railtrack’s operational requirements.   The system will implement cab signalling which conforms to ETCS level 3.  Automatic Train Protection (ATP) will be an integral feature.  Signalling information will be communicated to trains via a GSM-R radio network.

This form of Train Control System will set the pattern of systems in the 21st century

It will use a Global System Mobile Communications – Railway (GSM-R) radio network which allow constant data communication from a controlling system to the moving train.  GSM-R is the Railway variant of the system widely used by digital-cellular mobile phones.  The TCS will also implement variable moving block signalling on all lines, and it will be possible to run trains in both directions an any line up to 250km/h, and give a minimum headway of 90 seconds for following trains when running at that speed.  No line-side signals will be provided (except possibly at a small number of stations where non fitted trains use the WCML for a short distance) and all information to the driver will be displayed within the driving cab.  ATP will also be provided.

The above global statement describes the high level requirement for any future TCS but of course other subsystems will need to be developed such as:

Most of the above are caused by a requirement to remove from the trackside infrastructure both personnel safety reasons and for better reliability. The major system falling into this category is the humble track circuit and its insulated block joints.  Whilst it is recognised that some track circuits will always be needed, it is anticipated that there will only form about 0.5% of the existing requirement left.

The other needs of the new system will obviously follow in principle the requirements of any train control system.

 There will still be a need for an interlocking that will:

Each train will inform the interlocking of its location and length by means of the GSM-R radio network and using its received movement authority to provide automatic protection against exceeding that permission.

Other enhancements will be provided such as staff warning systems to alert trackside workers, and include the ability to grant track possessions directly via the data radio.

Needless to say, the new system will also have to control level crossings and allow the imposition of emergency and temporary speed restrictions.  All of these functions of course being implemented in a safe and consistent manner.

As the future TCS is implementing Dynamic Moving Block, the movement authorities will be infinitely variable taking into account parameters such as location, train length, braking performance, routes, permitted train speeds as well as geographical route features.

Safety and reliability are expected to be better than existing systems on a railway operating 24 hours a day, every day.

Availability to the system is expected to be high.  The mean time between Train Control System features which cause a delay to a train greater than 20 minutes is expected to be 10 days, and maintainability (MTTR) is to be less than 15 minutes on-site.

All these requirements of course mean that the system will be configured with a high level of redundancy including of course the radio and train born systems.  The advantage of modern computer systems also allow a very high level of diagnostic aids and data storage, plus the ability to communicate this information to a central monitoring point.

It is hoped that the New Train Control System for the WCML will be tested and approved for use by the year 2000.  (Please see brief update at the top of this article)

Management Centres (Formerly Control Centres):

It would be fair to say that the development of railway control centres has been tightly coupled with the implementation of the signalling systems.  The scope of these control centres has been constrained by the needs of local operation and has not been extended to consider the opportunities for strategic and tactical management developments or the optimisation of the roles and responsibilities of the personnel contained therein.  Their development has been slow to respond to the potential of systems and support tools now available.

Within the United Kingdom the railways have been reorganised and divided into separate business units.  For one unit, the operators of the infrastructure, their remit is to supply train paths to their best commercial advantage.  However, for it to be a realistic and meaningful operation, it must be done in co-operation with all of the other units and suppliers of services.  This co-operation will need better management systems using more accurate and reliable information on train and infrastructure operation, hence the change of name from Control to Management Centre, which best describes the function of the future, certainly within Railtrack plc, the owner of Britain’s infrastructure, a Management Centre should:

Currently, there is a clear demarcation between operational functions (signalling zone control, electrification, production management etc.).  With this demarcation often goes physical separation.  This separation requires the use of telephone and fax facilities, with little face-to-face communication.  Because of this, misunderstanding can occur especially in complex, often paper based procedures, which in themselves are not interlocked with the systems they are controlling.

Forecasting tools are not readily available, although some effort is now being made to use off-the-shelf packages linked into bespoke systems to provide some performance feedback.

There is currently no condition monitoring information coming up from the ground that can be used to forecast and instigate repairs and maintenance requirements.  This is going to be a vital component in future systems and provision will need to be made for such systems, to avoid them becoming just another ‘bolt-on’ facility alongside other tools.

The underlying problem with existing railway control and management systems is that they are reactive and not pre-emptive.

In addition to the equipment and process deficiencies, and because of the separation of operational functions, there are numerous buildings through-out the region of the infrastructure that incur a considerable overhead in manning the maintenance and high management overheads.

What Features do we need to Incorporate into a Management Centre?

Before any work is undertaken in defining new systems for inclusion in a Management Centre, there is a need for the re-evaluation of the business processes and operations of a railway organisation.  Most railways around the world continue with the separation of signalling and electrical control operators, operational controllers and production management staff.  The operational needs of the infrastructure must first be identified and then the roles and responsibility of personnel can be defined.  In the future the functions could be integrated into a single job specification for an ‘Infrastructure Operator’, which is far removed from the current divisions.

Having identified the processes required for the operation of a railway, attention then needs to be directed towards the quality of the operational and tactical systems that are implemented to assist those processes.  The life-blood of the system is the quality and quantity of information that is made available to the operator.  Systems can be implemented for the automation of the more mundane, repetitive and forecastable activities (such as possession management, isolations, bridge bashes etc.), leaving the operator to deal with the wider regulation and management issues.  The assistance provided by the systems with the repetitive activities will also increase the consistency of the approach, giving a more reliable and forecastable result that is less prone to effects of individual competencies.

With the greater use of systems and the potential reduction in manpower, there is the opportunity to bring together all of the personnel required to operate a region of infrastructure.  This collocation will increase the reliability of communication streams as more contact would be face-to-face, with a smaller number of interfaces and media.

One of the deficiencies of the current systems is their reactive nature.  To combat this, greater use needs to be made of trend analysis and forecasting tools.  With the additional, higher quality information that will be available, background forecasting tools will be able to ‘fast-run’ the current timetable and identify possible route contentions, supply deficiencies or rolling stock/staff short-falls.  With the necessary advanced warning the operator can turn his attention to avoiding the difficulty, rather than dealing with it after the event!

It is important that the systems developed for the Management Centre are modular - but integrated!  The development of support tools will become an increasingly dynamic activity.  The operators will be continually identifying deficiencies in their systems.  They will be able to specify how they can improve on the tools they have been given.  Only if the systems are modular by design, with clear, self-contained functionality and interfaces can the systems be developed and deployed that will enhance the operators function in an appropriate timescale.

In addition to the operational benefits from better information systems, better forecasting and planning will improve the utilisation of maintenance and renewals and the development of optimum train path timetables.  In other words, the Management Control Centre of the future is an integrated Management Tool that is used to increase sales, reduce costs and therefore, increase profits.

Conclusion:

The start of the 21st century is shortly arriving!  Signalling and Telecommunications for Railway operation is going to radically change.  There is a political and demographical need around the world to enhance the capacities and qualities of our transport systems, and notwithstanding research into alternatives, the steel wheel on a steel rail will still be the most economical transport system for many decades for urban and interurban travel.  Dynamic Train Control Systems such as ATCS and ETCS will, I believe, show themselves as being a practical and cost effective means of achieving a rebirth of our railway systems.  Certainly in Britain I fully expect the Control of the Railway network to be executed from a ‘few’ management centres within the next 15 to 20 years.  They will in many cases interface to existing signalling systems, but in time, with the installation of transmission based Train Control Systems, the majority of Management Centres will be interfacing to such systems.  They will then become the fully integrated Management System for the rail business, not only controlling trains, but also receiving and disseminating condition information on rolling stock and the state of the infrastructure.