Providing accurate, understandable and timely information to the travelling public is an ever present challenge. Described once as not a can of worms, more a bucket of snakes, this is a fairly apt description for the task. Anyone who thinks it is easy has no concept as to what is required or how it should be done. The recent hard winter with large amounts of snow demonstrated the problem all too clearly. Train services were inevitably disrupted and particularly in the politically-sensitive electrified third rail areas of the former Southern Region. Here the massive build-up of snow and ice on the conductor rail made it a huge challenge just to keep trains moving, let alone providing accurate information as to which services would be operating. It is rarely as bad as this, but even minor disruption can present a significant challenge to the flow of train running information.
So why is it so difficult and what can be done to improve things? the rail engineer talked with Chris Scoggins, the Chief Executive of National Rail Enquiries, part of the ATOC sphere of operations, to find out.
Old and Emerging Requirements
The occasional train traveller usually associates passenger information as something that he / she acquires from the Internet to look up train times and fares prior to making the journey and at the station to get confirmation of train running and platform. These two elements remain vitally important and the accuracy of the information at this stage can have a significant impact on the perception of rail travel. The classic engineering terminology of right and wrong side failures just do not apply in this scenario. Wrong information given out is as damaging as not having any information at all; indeed some would argue that displaying wrong messages is worse than showing nothing.
The seasoned traveller, however, increasingly wants much more than this. The advent of mobile phone networks and on-air data provision has meant that personalised information for the planned journey should be capable of being given out both prior to journey commencement and en route. Any disruption that would cause the journey to be modified should be advised by text message or email direct to the person’s mobile device, be it laptop, iPhone, Blackberry or mobile telephone. With this upping of the facilities comes the need for even greater accuracy so that the intending traveller can modify their day’s schedule with confidence.
The railway operates to a timetable and from this it should be possible to derive data that is capable of being fed to all users who promote and publicise train travel. However, as in all modes of transport, the delay and disruption that can occur will cause the timetable to be deviated from in both planned and unplanned situations. It is during these times that information provision is at its most important and it has been a challenge for the rail industry over many years to get accurate updates to the running of services collated and distributed to those that need to know.
Enter the Darwin concept – an initiative and development by the Train Companies, ATOC and Network Rail to get much greater accuracy in the compilation of train information data and the distribution of it to a widening user community. Described as a Real Time Train Prediction system, Darwin draws data from a number of sources, assesses the information and then intelligently predicts what this will mean to the ongoing train service. The sources are:
- Integrated Train Planning System (ITPS) – this is the Network Rail basic timetable, which is compiled twice a year but updated every 24 hours & distributed every night to all rail companies and external bodies that require timetable information
- TRUST (Train Reporting Using System TOPS) – a system that logs train movements and timing at selected passing points on the network, distributing the report to train and network control offices
- Train Describers (TDs) – the part of the signalling system that informs signallers of the identity and whereabouts of every train on the control panel with all berth steps (real time train movements) being available as a data message
- Control Room Information Controllers – known as the Tyrell system that is used to provide structured messages to TOC staff on cancellations, short train formations, etc.
- Some CIS control desks where train departure updates are decided locally
- Darwin Workstations – provided in the National Rail Communications Centre (NRCC) and in TOC control offices where direct input to the system can be made.
All these information packages are sent to Darwin as they happen so a high number of data messages are constantly being received.
What does Darwin Do?
The success of any traveller information system will lie in its ability to predict the future to a high degree of accuracy. The algorithms of Darwin are designed to:
- Store the basic timetable data plus planned short term amendments
- Receive train running data from various sources to be able to predict how the train service is operating in real time
- Identify trains that are not running to schedule
- Compile the necessary data to produce amended train running information along the railway geography
- Send this data to passenger information communications distributors provided by various third parties, including over 30 mobile phone companies, who will make this available to customers.
A typical situation could be the imposition of a temporary speed restriction maybe because of extreme heat or high winds. The system must assess how any reduction of speed will affect train running times, not just for a single train, but for all trains using that route. In the predictions made, train timings need to be marginally optimistic so as to ensure that passengers get to the departure point before the actual train arrival, thus perhaps dissuading people that they have time for the final cup of coffee!
Darwin is a new system that was introduced in 2009 but it had two predecessor systems that first went live in 2003. Darwin and its predecessors recognised the need for a national system by which customers can access the best possible real time running information for all trains nationally, using all the normal, easy to use, customer contact channels. Initially the service was offered only as stand alone information but is now built into journey planners as well. Darwin is constantly evolving (hence its name!) and is now at Generation 3.
The Darwin Architecture and Supply Base
The main Darwin contract is with Thales Group who, as well as doing the development work, also host the system, manage the provision of the service and maintain the entire Darwin architecture. The work is done from their Stockport premises. However, National Rail Enquiries own the intellectual property rights of the system and the programming code. Other specialist firms are used by Thales when the need arises. Such is the pace of change that a new release is being issued every 4 months.
Darwin runs in a live/live configuration from two data centres in the north of England, each site being a duplicate of the other and linked together by different commercial telecom providers. Should one site fail, the other has the capacity to operate the entire system. The NRCC (National Rail Communication Centre), located at Doncaster, is responsible for monitoring the quality of data within Darwin. They can give valuable assistance to the TOC control rooms should they need help in keeping Darwin up to date with high levels of operational decisions during severe disruption.
Impact on Station Based Information Displays
Automated provision of passenger information systems at stations has been around for more than 30 years but often suffered from the accuracy of the data that was used to drive the displays. The former Southern Region of BR pioneered a system to actuate displays and announcements from a timetable data source but it was found necessary to have a dedicated team at the Waterloo HQ to amend the data so as to cover special workings, weekend engineering work, diversions and platform changes. Gradually this system was expanded to all busy areas of the national rail network and was improved over time by the many suppliers now offering products in this field. However, if disruption occurs and the planned timetable deteriorates, the system quickly fails to cope, with the result that much misleading (or even wrong) information is posted to displays causing mild humour at best and ridicule at worst.
Using Darwin data to provide a real time updating of station information systems was a natural progression and a trial has recently started at 17 Virgin stations on the West Coast Main Line, including major interchange points such as Crewe, Preston and Birmingham International. This has been done in co-operation with Amey who were the original providers of the station CIS equipment and who link the station systems together with an independent data network. It is not the intention that any station CIS system connected to the Darwin data source would need a hardware upgrade. Some modification to the data provision routines will be necessary, which Thales will provide as part of their Darwin contract. The Virgin stations trial has been operational for nearly 2 months and it is calculated that the accuracy of the displayed information has improved to 99%.
Usage and Future Plans
Statistics for general rail enquiries are illuminating. In 2002 there were 62 million train enquiries made to National Rail Enquiries primarily by telephone. In 2010, there were 250 million but less than 5% of these was by phone. The increasing reliance on data sources tells its own tale. Information has therefore to be provided to cater for business and public data accessing and more than 180 licences have been granted for the receipt of Darwin outputs. 30 of these are mobile phone suppliers – used mainly for iPhone, Android and Blackberry customers – with a small charge being made for every user application. An average user accesses the system twice a day, mainly to check how a particular train service is running. Web sites such as Twitter and broadcasters (BBC and ITV) also receive the data for onward transmission as do some travel agents. More licences are being granted all the time.
Rolling out Darwin data to station CIS systems nationally is a longer term project. A phase 2 rollout to 1900 stations is seeking funding later this year with a 2 to 3 year implementation. This will include the major stations that are managed by Network Rail. Getting real time train positioning data is still a problem on lines not provided with train describers and equipping trains with GPS receivers is being investigated as a possible solution.
Darwin is seen as a major step forward in both the quantity and accuracy of train running information on offer for public consumption. It will interface with various types of communication media as well as improving the displayed information at stations. At a time when certain sections of the media find rail transport an easy target for criticism, the work being done by the industry through Darwin will go a long way to dispelling the perception that train service information is often inadequate.