Where’s My Tube
When in any big city around the world, the chances are you will ride on its metro or underground system. But knowing which line to travel on and the station to get off at can be a perplexing experience, especially if you are not used to this mode of travel. Much will depend on the information you are supplied with, be it paper maps, signage at the station, or indicator systems once you are on the platform.
London is probably the most visited city in the world and most tourists, as well as the local inhabitants, will use the underground system on a daily basis. Getting meaningful and accurate information out to passengers is a vital part of the train service offering and has been ever since London Underground began operations in the Victorian era. A recent talk given to the IRSE London & South East section informed the audience of the history of passenger information and how modern day IT and train tracking systems bring sophistication and accuracy to the display screens.
A history of displays
London Underground (LU) has provided platform displays for over a century. These may be summarised as follows:
- Indicator boxes with all destinations printed on enamel plates. An adjacent light box illuminated an arrow to indicate the first train. At some stations and where a train describer system existed, the sequence of approaching trains would be shown by illuminating 1, 2, 3. The system dates from 1906 but only one remains today at Earls Court where it is a listed feature.
- Light boxes. Train destinations printed on transparent glass compartmentalised within a box. Once the train destination is known, the correct compartment is lit up.
- First train indicator. These date from the 1960s but only show the destination of the next train and all are now removed.
- Two train indicator. An improvement on the former showing the destination of the first and second trains but nothing else.
- Dot matrix displays. These are now installed at virtually every station. The system dates from 1981 and was a 17-stage project to provide indicators at 340 platforms. Using Light Emitting Diode (LED) displays, not only do they show the destination of the first, second, and third train, with sometimes a fourth train, they incorporate a countdown feature that informs the time in minutes for the arrival of each of the trains displayed. Additional safety information is also displayed – see picture.
- At important interchange stations, displays may be provided at locations between the concourse and the platforms to indicate the times of the next trains from each of the platforms. An example is at Euston where the Northern Line is served by two different branches with trains to the same destination (e.g. Edgware) being from two different platforms. The traveller then knows from which platform to go to for the first train.When first introduced, an average station cost around £12,000 for the indicator provision, nowadays this is typically £34,000. On some lines, e.g. the Metropolitan northwards from Baker Street, some trains to the most northerly suburbs do not stop at all stations. The stopping pattern is indicated for the first train using a second line of text.
- Summary displays. Sometimes referred to as Rainbow Boards, these are positioned in the concourse areas of LU stations that typically serve the main line termini. They show the service level on all the LU lines to indicate whether any delays or line closures are taking place at the current time. Originally these showed the level of service on every line but the display has been simplified to indicate lines where disruption is occurring with an added comment ‘good service on all other lines’. Passengers entering the underground system can thus make decisions as to the route of their ongoing journey if their intended route is disrupted.
Acquiring and assessing the information
Key to obtaining and distributing train running information is the train describer system (TD) within the signalling system. This has a long history, dating back over 100 years. Each train has a description which is unique for a particular journey made up of a five-letter code (ABCDE) or combinations within that. Examples for the District Line are:
- AD: Plaistow to Richmond.
- AB: Whitechapel to Northfields.
- C: Tower Hill to Putney Bridge.
Originally, the data was stored on an electromechanical drum but, from the 1960s, a 10-store system using individual circuit boards has been deployed. In time, this too has been replaced by computerised systems on the Central, Jubilee, Victoria, and Northern lines as their signalling system was replaced with Communications Based Train Control (CBTC) technology. The latest upgrade is on the 4LM project (Metropolitan, District, Circle, Hammersmith & City lines) where the Thales Seltrac CBTC system is being deployed.
The train description is then linked to the track circuits within the signalling system, and as each track circuit is occupied by the progression of a train, so the TD steps forward. As lines have been modernised, track circuits are being replaced by other forms of train detection, but the same progressive stepping of trains is maintained. Primarily, the description is used by the signallers in the control rooms who have a VDU line diagram display showing the whereabouts of every train. Correct routing is usually achieved automatically but the signallers can change this if need be. The train description has a secondary function to feed the passenger information displays, which has been the situation over many years and remains largely true today.
TD systems have been computer based since the 1980s thus allowing greater intelligence to be gathered which has helped the accuracy of the platform display information. With the introduction of countdown minutes, the challenge has been to provide the timing data with the appropriate degree of accuracy. LU approached this by establishing timings from nominated stations. An example would be Mansion House to St James’s Park, some five stations away. The time can of course vary depending on whether or not the train is delayed because of busy platforms and extended dwell times. When first introduced it led to the humorous differentiation between real minutes and LU minutes, which were invariably longer!
All of this intelligence is within the customer information system that can interpret train describer codes and compute the actual information to be displayed. It can get it wrong, and instances of the wrong destination being displayed do occur. When this happens, the station public address system is used to advise passengers to check the destination on the front of the train. At busy stations where platforms are staffed, the platform announcer will give the necessary correction and advice over the PA system.
Information for all
While the quality, production and accuracy of information has been measurably improved, so far this article has confined itself to station and platform displays as well as improved data for signallers and controllers. But what about travellers who have not yet reached the station, or indeed LU staff in depots or offices?
So has emerged Trackernet which was initially designed to provide train running information to any LU member of staff via an internal Intranet data system on their business PCs, including remote access from home. Particularly important was having this information at depots and stations where crews book on and off duty. This brought into focus the distinction between Operational Technology (OT) and Information Technology (IT) and the need for cyber security. The system was modernised for the Central Line upgrade with better graphics and signal displays, and caters for Automatic Train Operation (ATO).
Trackernet is deployed wherever a line has a centralised control system, but it is also capable of capturing information from non-computer-controlled lines e.g., the Piccadilly branch to Uxbridge. A more recent version has been developed for the 4LM routes which includes train timing information. It even has a link to control the light boxes at Earls Court. As well as computer terminals, the information is extended to give visibility on all TfL managed mobile devices such as iPads via the station wifi networks or increasingly the 4G radio networks. None of this type of information is safety critical, which eases the design criteria.
Extending Trackernet data to inform the general public was an obvious next step. This is achieved via the TfL App which also shows station maps and locality with pictures of station entrances and platforms. The intent is to give times for incoming trains so that if a train is very crowded, you will know when the next one will be coming. A half hour prediction window is envisaged.
What of the future?
Improvements to train running customer information will always be ongoing. In the immediate future, signalling derived prediction lists will be extended to more areas together with more information about train details such as individual car numbers for internal decision making. Train loading information using either pressure in the air suspension system or static deflection of the springs, the data from which will transfer to the stations’ indicator systems, will be a natural progression to indicate platform positioning for passengers to the less crowded part of the train. Thameslink and Elizabeth line trains already have train loading data but is only shown within the carriages and not on their platforms. Integration with other transport systems – buses, cabs, main line rail – is another vision for the future.
London Underground has come a long way since its original indicator systems were installed and the information provided currently is amongst the best in the world. As in any metro system, when the service is running to its timetabled pattern, the information being provided is relatively straightforward; it is when things go wrong that the system is put to the challenge.
Thanks are expressed to Adam Ladds and Andy Wiseman who gave the original talk to which has been added additional information from a variety of sources.
Image credit: TfL