Automation of the New York City Subway
The New York Metropolitan Transportation Authority (MTA) operates the New York City Subway, which is mostly manually operated. The subway system currently uses Automatic Block Signaling with fixed wayside signals and automatic train stops. Many portions of the signaling system were installed between the 1930s and 1960s. Because of the age of the subway system, some replacement parts must be custom built for the MTA, as they are otherwise unavailable from signaling suppliers. Additionally, some subway services have reached their train capacity limits and cannot operate extra trains with the current Automatic Block Signaling system.
The MTA has plans to upgrade the entire New York City Subway system with communication-based train control (CBTC) technology, which will control the speed and starting and stopping of subway trains. The CBTC system is mostly automated and uses a moving signaling system – which reduces headways between trains, increases train frequencies and capacities, and relays the trains' positions to a control room – rather than a fixed position signaling system. This will require new rolling stock to be built for the subway system, as only newer trains can use CBTC systems.
- 1 Current signaling system
- 2 Automation overview
- 3 History of New York City Subway automation
- 4 References
- 5 External links
Current signaling system
New York City Subway generally distinguishes its current signals into:
- automatic signals, controlled only by train movements
- approach signals, like automatic signals, can be forced to switch to stop aspect by interlocking tower
- home signals, route set by interlocking tower
- additional signals (call-on, dwarf, marker, sign, time signals)
Common automatic and approach signals consist of one signal head showing one of the following signal aspects:
- stop (one red light); with special rules for call-on and timer signals
- clear, next signal at clear or caution (one green light)
- proceed with caution, be prepared to stop at next signal (one yellow light)
Where different directions are possible, the subway uses both speed and route signalling:
- upper signal head for speeds
- lower signal head for routes (with main route shown green and diverging route shown yellow)
The system currently uses block signalling, which is used in other systems such as the Toronto subway and RT. The block signals that the New York City Subway currently uses is identical to those on the RT's signaling system.
Stop. Passing this signal trips the train stop
The system also has automatic and manual key-by red lights. They involve the operation of an automatic stop with an automatic or manual release, then a procedure with caution, with preparations to stop in case of debris or other obstructions on the track.
A Station Time signal, also used in the system, allows trains to close in on each other in a station if they slow down enough.
Aside from some parts of the original IRT system, the entire subway uses this signalling system. In some of the IRT lines, the lights, from bottom to top, are yellow, red, and green. On the rest of the system, the lights from top to bottom, are red, yellow, and green.
Interlocking signals are used in interlockings, which are any areas where train movements may conflict with each other. They are controlled by human operators in a signal tower near the switches, not by the trains themselves. A train operator must use a punch box to notify the switch operator of which track the train needs to go. The operator has a switchboard in their tower that allows them to change the switches.
Interlocking signals also tell switch operators which way switches on the subway are set. The following interlocking signals are used on the New York City Subway:
Timed block, timer has not yet run out, next block is timed as well as lunar aspect is indicated (in this example this signal would only clear to yellow over green)
To precisely specify locations along the New York City Subway lines, a chainage system is used. It measures distances from a fixed point, called chaining zero, following the twists and turns of the railroad line, so that the distance described is understood to be the "railroad distance," not the distance by the most direct route ("as the crow flies"). This chaining system differs from the milepost or mileage system. The New York City Subway system differs from other railroad chaining systems in that it uses the engineer's chain of 100 feet (30.48 m) rather than the surveyor's chain of 66 feet (20.1168 m). Chaining is used in the New York City Subway system in conjunction with train radios, in order to ascertain a train's location on a given line.
Trains using CBTC interact with the locate themselves based on measuring their distance past fixed transponders installed between the rails. Trains equipped with CBTC have a transponder interrogator antenna beneath each carriage, which communicates with the fixed trackside transponders and report the trains' location to a wayside Zone Controller via radio. Then, the Controller issues Movement Authorities to the trains. This technology upgrade will allow trains to be operated at closer distances (slightly increasing capacity) and will allow the MTA to keep track of trains in real time and provide more information to the public regarding train arrivals and delays. The trains are also equipped with high-tech computers inside the cab so that the conductor could monitor the train's speed and relative location.
Only newer-generation rolling stock that were first delivered in the early 2000s—the R143s and 64 R160s (8313-8376)—are equipped for CBTC operation. Future car orders, specifically the R179, the R188, and the R211, will also be designed to be CBTC compatible. After the retirement of the R68 and R68A cars, all revenue cars, except those on the G, J, M, Z and S trains, will be equipped with CBTC. The BMT Canarsie Line was the first line to implement the automated technology, using Siemens's Trainguard MT CBTC system.
Most subway services are already at capacity, in terms of train spacing, during rush hours, except for the 1, G, J/Z, L, and M trains (the L service already is automated with CBTC). Therefore, transit planners are viewing the installation of CBTC as a way to free up track capacity for more trains to run, and have shorter headways between trains. However, installing CBTC in the New York City Subway is harder than in other systems due to the subway's complexity. The MTA hopes to install 16 miles of CBTC-equipped tracks per year, while the Regional Plan Association wants the MTA to install CBTC signals on 21 miles of tracks per year.
It is of note that even without CBTC, the system is currently retrofitted to operate at frequencies of up to 60 trains per hour (tph) on the IND Queens Boulevard Line (30 tph on each of the local and express pairs of tracks made possible by the Jamaica – 179th Street terminal) and 33 tph on the IRT Flushing Line. The BMT Canarsie Line is limited to a 26 tph frequency due to the bumper blocks at both of its terminals; however, the IRT Lexington Avenue Line operates at frequencies of 27 tph without CBTC. By contrast, lines on the Moscow Metro can operate at frequencies of up to 40 tph, since lines in the Moscow Metro, unlike in the New York City Subway, typically have four sidings past the terminals instead of bumper blocks or one or two sidings.
History of New York City Subway automation
42nd Street Shuttle automation
The 42nd Street Shuttle, which runs from Grand Central to Times Square, was briefly automated from 1959 to 1964. The chairman of the Board of Transportation, Sidney H. Bingham, in 1954, first proposed of a conveyor belt like system for the shuttle line. Charles Patterson, a few years later, as the President of the newly formed New York City Transit Authority (NYCTA) told of a vision of automated mass transit, without relying on the use of motormen. General Electric responded to Patterson's speech, stating that this technology was feasible, and that the company was interested in the idea of automating the New York City subway.
The idea of automation at that time relied on commands that were sent to the train while the train is at a station, to keep its doors open. When the commands cease, the doors would promptly close. A new series of commands would start the train and gradually accelerate it to 30 miles per hour (48 km/h), maintaining that speed. This is only under the condition that no other command overrides it. When approaching the next station, there was an insulated rail joint, where if the train had passed it, new command would come to slow it to 6 miles per hour (9.7 km/h). Inside the station, new commands at another insulated rail joint would command the train to stop. At the station, the train would have opened its doors, reversed course (as this is a two station shuttle line) and the lighting for the directional signs would be changed to match its new destination.
Sea Beach Line test track
Representatives of General Electric, Westinghouse (traction), General Railway Signal (GRS) and Union Switch and Signal (US&S) (signals) and WABCO (Westinghouse Air Brake Company - brakes) met with Patterson and together planned to automate the 42nd Street Shuttle as a prototype for an automated system. The NYCTA was to supply the rolling stock, while the signal companies were tasked with the installation, maintenance and technological oversight of the automation process, including signalling. An express track on the BMT Sea Beach Line was first used to demonstrate the technology, before it could be applied for passenger service. The stretch of track from 18th Avenue and New Utrecht Avenue was used, as it best replicated the length of the shuttle line.
A handful of R22s were used for the line. The cars, however, were fitted with different types of brake shoes, to see which one would negotiate the rail joints better. It was eventually found that the automated trip took 10 seconds longer than manual operation (about 95 seconds, compared to 85 seconds). As the tests on the Sea Beach line progressed, grade time stops were added to ensure safety on the line, and on the 42nd Street line. The train was dubbed SAM, and was to operate on Track 4 of the shuttle line. It was demonstrated to officials in 1960, and was still running without passengers until January 4, 1962. A motorman was to be present and take over in case if there were any problems.
The demise of the line came with a fire at Grand Central Station on April 21, 1964. The fire was not related to the automated trains. The automation, however, provided the framework for automated rapid transit technology on BART (San Francisco) and PATCO Speedline (Philadelphia-Camden-Lindenwold).
After the fire that destroyed the automated shuttle subway cars, ideas for automation lay dormant for years, until an intoxicated motorman caused a train crash at Union Square station that killed 5 people and injured 215. The collision was a catalyst to a 1994 business case outlining arguments for automatic train operation and communications-based train control, which led to the automation of the BMT Canarsie Line starting in the early 2000s.
Canarsie Line CBTC
The Canarsie Line, on which the L subway service runs, was chosen for CBTC pilot testing because it is a self-contained line that does not operate in conjunction with other subway lines in the New York City subway system. The 10-mile length of the Canarsie Line is also shorter than the majority of other subway lines. As a result, the signaling requirements and complexity of implementing CBTC are easier to install and test than the more complicated subway lines that have junctions and share trackage with other lines.
The project was first proposed in 1992 and approved by the MTA in 1997. Installation of the signal system was begun in 2000 and was mostly completed by December 2006. Due to an unexpected ridership increase on the Canarsie Line, the MTA ordered more cars and these were put into service in 2010. This enabled the agency to operate up to 26 trains per hour up from the May 2007 service level of 15 trains per hour, an achievement that would not be possible without the CBTC technology.
Flushing Line CBTC
The next line to have CBTC installed will be the IRT Flushing Line and its new extension (7 <7> trains). The Flushing line is being chosen for the next implementation of CBTC because it is also a self-contained line with no direct connections to other subway lines currently in use. Funding is in the 2010–2014 capital budget for CBTC installation on the Flushing line, with scheduled installation completion in 2016. The R188 cars have been ordered to equip the line with compatible rolling stock. This order consists of new cars and retrofits of existing R142A cars for CBTC. CBTC, in conjunction with the 7 Subway Extension, will allow the 7 <7> services to run 2 more tph during peak hours (it currently runs 27 tph, but has a built-in capacity for 33 tph).
Queens Boulevard CBTC
The MTA is also seeking to implement CBTC on the IND Queens Boulevard Line. CBTC is to be installed on this line in five phases, with phase one (50th Street/8th Avenue and 47th–50th Streets – Rockefeller Center to Kew Gardens – Union Turnpike) being included in the 2010-2014 capital budget. The $205.8 million contract for the installment of phase one was awarded in 2015 to Siemens and Thales. Planning for phase one started in 2015, with major engineering work to follow in 2017. The total cost for the entire Queens Boulevard Line is estimated at over $900 million. 
The automation of the Queens Boulevard Line means that the E F services will be able to run 3 more trains during peak hours (it currently runs 29 tph). This will also increase capacity on the local tracks of the IND Queens Boulevard Line. However, as the line hosts several services, installation of CBTC on the line can be much harder than on the Flushing and Canarsie lines.
Culver Line CBTC test track
In addition, funding is allocated for the installation of CBTC equipment on one of the IND Culver Line express tracks between Fourth Avenue and Church Avenue. Total cost is $99.6 million, with $15 million coming from the 2005-2009 capital budget (phase one) and $84.6 million from the 2010-2014 capital budget (phase two). The installation is a joint venture between Siemens and Thales Group. The estimated completion date has been placed on March 2015; the installation is expected to be permanent. Should Culver Line express service be implemented, the express service will not use CBTC, and testing of CBTC on the express track will be limited to off-peak hours.
CBTC on other lines
The MTA projects that 355 miles of track will receive CBTC signals by 2029, including most of the IND, as well as the IRT Lexington Avenue Line and the BMT Broadway Line. The MTA also is planning to install CBTC equipment on the IND Crosstown Line, the BMT Fourth Avenue Line and the BMT Brighton Line before 2025.
However, the Regional Plan Association prioritizes the Lexington Avenue, Crosstown, Eighth Avenue, Fulton Street, Manhattan Bridge, Queens Boulevard, Rockaway, and Sixth Avenue subway lines as those in need of CBTC between 2015–24.
- Chan, Sewell (2005-01-14). "Subways Run by Computers Start on L Line This Summer". The New York Times. Retrieved 2007-05-24.
- Subway Signals: Approach, Automatic, and Marker Signals
- Subway signals: Interlocking
- JoeKorNer - New York City Subway Chaining
- CBTC: Communications-Based Train Control on YouTube
- Twenty-Year Capital Needs Assessment
- Neuman, William (2007-05-22). "For Less Crowding on L Train, Think 2010, Report Says". The New York Times. Retrieved 2007-05-24.
- RPA CBTC plan
- Mark S. Feinman, Peggy Darlington, Joe Brennan, David Pirmann. "IRT Times Square-Grand Central Shuttle." nycsubway.org. Accessed 2012-08-06.
- New York City Transit Authority (1960). Automated Train Brochure.
- McFadden, Robert D. (September 1, 1991). "Catastrophe Under Union Square; Crash on the Lexington IRT: Motorman's Run to Disaster". New York Times. Retrieved June 11, 2014.
- Pages 11–12
- http://www.mta.info/mta/news/books/docs/TYNA-Consolidated.pdf, pages 46-47
- chapter 2
|Wikimedia Commons has media related to New York City Subway signals.|