Minutes prepared by: Dr. Alan F. Rumsey

In introducing the meeting, Pete Furkey, PATH, noted that these are exciting times for the signaling industry and this CBTC Peer Group had a unique opportunity to shape the direction the industry will take well into the 21^st century.

Michael DePallo, Director/General Manager of PATH, welcomed the meeting participants, and noted that Communications-Based Train Control was a technology whose time has come. CBTC will change the way transit agencies do business, by reducing operating and maintenance costs, while also reducing capital costs and improving safety and reliability. There are certainly hurdles to overcome, and hopefully this CBTC Peer Group will contribute to overcoming the hesitation associated with new technology.

Alan Rumsey (Parsons/De Leuw, Cather) summarized the CBTC standardization efforts of IEEE Working Group #2 and described the initial mandate of the working group, progress achieved, problems encountered, proposed revisions to the group mandate, and required transit agency involvement.

The general objective of the IEEE standardization efforts is to minimize the acquisition costs of CBTC technology through standardization of key interfaces. Specifically, working group #2 had been attempting to establish standards for the train-to-wayside and wayside-to-train interfaces associated with basic train protection functions. Although good progress has been achieved in identifying similarities and key differences between the various CBTC systems currently available or under development, a number of commercial and technical difficulties had prevented suppliers from reaching a consensus on interface standards.

To build on the momentum established by the working group to date, it has been proposed that the group proceed with a revised mandate, focusing its attention less on the specific technical details of the train-to-wayside interfaces, and more on the overall performance and functional requirements/expectations for CBTC systems. Specifically, it has been proposed that working group #2 attempt to develop a standard Performance/Functional Specification for CBTC Systems. A Standard Specification for CBTC Systems should enable suppliers to focus their CBTC system development efforts more efficiently, enable transit agencies to streamline their CBTC system procurements, and minimize the amount of new design required (with associated costs and risks) for each new CBTC application. A matrix was tabled which attempted to summarize the performance/functional requirements that could be addressed by a Standard Specification.

Achieving such a standard would require a strong and consistent involvement of transit agencies on the Working Group, and transit agency interest in, and commitment to, such a process is now being sought. The IEEE full committee will be meeting on October 8, in Washington, to confirm the future direction of Working Group #2.

This topic was discussed further under agenda item 6, below.

Alan Rumsey and Geoff Hubbs (NYCT) summarized the status of the Canarsie Line Signal Modernization Project, New York City Transit’s Pilot Project for Communications-Based Train Control. The key project requirements and three phase procurement approach were described together with current project status, and planning for the upcoming proposal evaluation phase. The key project requirements are:

To establish a New Signaling Standard for NYCT, to permit future procurement of interoperable wayside and carborne CBTC subsystems from multiple suppliers

• To enhance safety of operations, through continuous overspeed protection

• To provide for maximum operational flexibility
• To provide for high system availability
• To prepare NYCT’s operating and maintenance staff, and
• To minimize disruption to service during implementation.

The procurement approach includes three phases. Phase I is a comprehensive demonstration test phase with three shortlisted suppliers, on the Culver test track. Phase II is the construction phase for the "Lead" contractor to re-signal the Canarsie line. Phase III is the Interoperability phase for two "Follower" contractors using Lead contractor Interoperability Interface Specifications.

With this approach, post-Canarsie, NYCT will have three (3) qualified suppliers of carborne CBTC equipment and three (3) qualified suppliers of wayside CBTC equipment. For future car procurements, carborne CBTC equipment will be included in the carbuilder’s scope of supply (using pre-qualified CBTC suppliers). CBTC wayside equipment will be procured on a line segment-by-line segment basis under the Signal Modernization program. Interoperability will be assured through interface specifications developed during the Canarsie Line pilot project. Key project milestones are as follows:

 The proposal evaluation approach will place an initial emphasis on the Proposer’s Qualifications and Technical Proposals for CBTC Systems. Final Price Proposals will only be negotiated with the three shortlisted contractors.

 NYCT have selected a number of qualified consultant teams to propose on their Independent Safety Consultant RFP.

Victor Grappone (LIRR) advised that the Long Island Rail Road’s (LIRR’s) signal strategy for the next 20 years was heavily dependent on CBTC, and that LIRR was looking to CBTC to deliver on its promise of reduced costs and increased capacity with, in addition, an ability to improve grade crossing control (through constant warning times and near side station stop control). LIRR expect to spend $600-$700 million in signal modernization in the next 20 years and are planning for all 1,000 cars to ultimately be equipped with CBTC. As retrofit costs are high, LIRR’s CBTC implementation will be tied in with new car procurements.

Three pilot projects are being planned:

• A grade crossing pilot project with a tie in to an Intelligent Transportation System (ITS) - to provide effective preemption of traffic signals - is currently being developed at New Hyde Park. GRS are the contractor for this project and a functional demonstration (not safety certified) is planned for late 1998. Safety verification of the system is expected to be complete approximately 12 months later, with active control of the three grade crossings. Between 3 and 6 trainsets will be equipped.
• LIRR have made available some limited funding to participate in Phase I of the NYCT Canarsie Line project, and would be prepared to accept the NYCT leader/followers as pre-qualified suppliers for LIRR if a good fit can be demonstrated between the systems being proposed for NYCT and that required by LIRR. To this end, during the Canarsie Phase I demonstration test phase, the shortlisted suppliers will be provided with an opportunity to demonstrate LIRR-specific functions (i.e. grade crossing operation, GPS-based train positioning), and submit a proposal for the implementation of interlocking functions without track circuits. The supplier’s compensation for Phase I will be increased accordingly. 
• LIRR are also planning a demonstration of basic train separation functions between Babylon to Patchogue which consists of approximately 13 miles of double track and approximately 3 miles of single track operation. Six to ten trains will be equipped with CBTC for this demonstration which will include a continuous demonstration of equipment operation in "shadow" mode with specific train separation demonstrations at night under controlled conditions. An RFP for this project should be issued in approximately 2 months, for demonstrations sometime in 1999.

 4.3 BART

Steve Mullerheim (BART) described the status of BART’s AATC program.

 The Technical Baseline for the AATC system had previously been distributed to the Peer Group for information, and Steve Mullerheim confirmed that an association had now been finalized with Harmon with respect to their involvement in the ongoing AATC program. It is important for both BART and Harmon that the AATC system is not an "orphan" system, particularly with respect to the radio hardware and network software.

 Phase II of the AATC program, which consists of safety certified AATC systems installed on 10 cars, five miles of mainline track and at 2 mainline station (Fruitvale and Lake Merritt), is scheduled to be completed by the Fall of 1999.

 Steve Mullerheim suggested that for certain functional areas of the AATC system design there was a high degree of commonality between the BART application and likely applications at other transit agencies. In other functional areas, the degree of commonality would be less, as indicated below:

Steve Mullerheim also described the proposed basis for the AATC safety assurance plan, and discussed the verification and validation exercises and design methodologies appropriate for hardware and software-related items of both the Vehicle ATC and Station Computer subsystems of the AATC system. Verification and validation exercises include:

Specification-based design and testing

• Walk throughs
• Fast prototyping and desktop simulations
• Formal software inspections
• Lab subsystem tests
• Lab integrated system tests
• Controlled track tests, and
• Analyses

Design methodologies include:

• Specification-based design and testing
• Software design standards
• Message encapsulation
• Numerical assurance
• Pre and post assertions
• Checked redundancy, and
• Validity checks

Finally, Steve Mullerheim suggested that the two biggest cost drivers in any CBTC system are the radio network design and safety validation, and hence priority should be given to these areas in any standardization initiatives.

Patty DeVLieg (MUNI) summarized the status of the MUNI ATCS project. The primary objectives of this project are:

• To obtain optimum throughput
• To improve safety
• To increase reliability and availability, and
• To reduce maintenance
• The ATCS design is required to support:
• Mixed fleet operations (SLRV and LRV2)
• Dual mode operations (ATCS with existing signaling system)
• "Open loop" transition from "street" running to ATCS
• Interfaces with existing signaling systems, and
• Installation & cutover without affecting revenue service

Limited ATCS operation (shuttle service) is currently being demonstrated and as a minimum, this shuttle service is scheduled to be in revenue operation by January 10, 1998. System integration issues, and integration with other related projects, have been major factors in the ongoing project delays. Design issues related to transitions into/out of ATCS territory have also proved difficult to resolve.

Peter Furkey (PATH) summarized the history of the PATH system from its beginnings in 1908 through to the current system configuration, with safety of train movements based on electropneumatic train stops and wayside signals. PATH does most of its signal design work in-house.

 The existing signal system is approximately 30 years old, and is very similar to the existing NYCT wayside signal/trainstop technology. PATH are planning to replace the existing signal system, and move to automatic train operations, in conjunction with new car procurements. The present signal system has large number of timed signals to enforce speed in curves; much fewer track circuits would be required with either a coded AF track circuit solution, or CBTC.

 For PATH, the objectives/advantages of ATO are:

• Maximizing throughput
• Centralization of equipment
• Elimination of track clutter
• Enhanced protection for track works (less need to be on track)
• Accuracy of programmed station stop
• Uniformity of operation, and
• Reverse running with following moves.

PATH plan to issue an RFP for consultant services, to put together the contract documents, later this year or early next year. One of the consultant tasks will involve a substantial amount of performance analyses, of the PATH-specific operations, to compare the operational benefits of an AF track circuit based ATO solution and CBTC. PATH would like to make the decision between an AF track circuit based solution and CBTC as soon as possible. With respect to carborne CBTC equipment, however, it was noted that the equipment required for CBTC was virtually the same as that required for a more conventional AF track-circuit based cab signaling/ATO package, with radios replacing the track circuit code receiver coils. With a CBTC solution, PATH would plan to retain track circuits for broken rail detection, switch locking and fall-back operations.

An RFP for new car procurement is to be let in 1998 with car delivery to be completed by 2002. PATH cars could be same as NYCT IRT cars.

An RFP for the signal replacement is planned for 2003, with the new signal system to be in service by 2006. An additional year is planned to demolish/remove the existing signal system.

John LaForce (SEPTA) noted that SEPTA is multi-modal, with commuter rail, heavy rail and light rail systems.

SEPTA’s heavy rail subway lines are being modernized with AF cab-signals and one person operation. There are no plans for automatic train operations (ATO). SEPTA do not expect to implement any more relay-based interlockings; vital processor-based interlockings will be used in the future.

SEPTA are looking at CBTC to enhance safety on the City Light Rail Subway Surface Trolley Tunnel. Approximately 80 2-block, three aspect color signals are currently installed in the 2 ½ mile tunnel, but with no enforcement. Rear end collisions and derailments have occurred in the past.

John Laforce advised that SEPTA hope to announce in approximately one month their procurement strategy for implementing CBTC in the tunnel.

SEPTA are looking for a "low risk" CBTC solution, with no ATS and no ATO related functionality. The primary objectives for the CBTC system are safe train separation assurance and civil speed enforcement. Approximately 114 vehicles will need to be equipped. SEPTA are looking more at a lossy coax, spread spectrum, RF solution rather than point-to-point free space antennas.

The CBTC system will need to support 30s to 1 min headways, with multiple berthing in stations. The CBTC system will interface to one existing interlocking. All of the existing track circuits will be removed, except at the interlocking. There will be no "fall back" system. In the event of CBTC equipment failures, vehicles will operate at an enforced restricted speed, on line-of-sight. There is no requirement for broken rail detection. With the light rail vehicles, broken rails have not been a problem at SEPTA. SEPTA would probably require broken rail detection on rapid transit lines, and would definitely require broken rail detection on commuter rail lines.

John Lewis (MBTA) advised that the MBTA were interested in CBTC, but had no firm plans to implement CBTC at this time.

Marty Lukes (WMATA) advised that WMATA currently have no plans for CBTC. New WMATA vehicles will have relay-based ATP systems.

Mohinder Singh Hare (NYCT), summarized the RF propagation testing that has been conducted to date at NYCT.

 In the last two years, NYCT has conducted a series of radio tests in collaboration with vendors to characterize the RF propagation within the NYCT environment, and to establish the feasibility of spread spectrum radios. Testing has been conducted with MATRA Transport, GRS/SENSIS Corporation, Harmon/Hughes Aircraft Company and GEC/Kasten Chase. Some of the conclusions include:

• A data link based on 2.4 GHz spread spectrum is viable.
• To establish a reliable/robust data link, detailed application engineering and associated testing will be required.
• Reliable transmission up to 2 miles above ground and 1.5 miles in single bore tunnels has been demonstrated
• Blocking train effect is significant; one possible solution is to use the train radios to relay messages.
• Antenna diversity can provide significant gain.

Alan Rumsey tabled the following examples of the types of issues that would need to be addressed in order to establish performance/functional standards for CBTC:

• How much variability in train types and train configurations should be supported by a "standard" CBTC system?
• Can a "standard" set of train operating modes, and mode transitions, be defined for CBTC systems. Is there a need to differentiate between light rail, heavy rail and commuter rail?
• Can a maximum operating speed be defined?
• Can it be assumed that a "standard" CBTC system will be controlled from a central control location, with local control provisions for failure management only, or should alternative line control philosophies be assumed? Should all ATS-level functions be considered optional?
• Different transit properties typically have different definitions for both "design headway" and "operating headway".
• Can standard headway definitions be developed for CBTC system applications?
• Should a quantitative safety criteria be defined for CBTC systems, if so, what should this criteria be (e.g. hazard probability of 10^-9 per system operating hour)? What "hazards" should be included in this calculation? How should the impact of a "failure" of safety software be addressed?
• Can a "standard" method be defined for specifying system availability/reliability requirements for CBTC systems? Can a "standard" method be defined for measuring compliance with these requirements?
• Can a "standard" method be defined for specifying system maintainability requirements or defining an overall maintenance philosophy for CBTC systems? Should "standard" requirements be established for CBTC system diagnostic provisions?
• Can "standard" environmental criteria be defined for CBTC equipment (ref. WG#8)?
• Can it be assumed that a "standard" CBTC system must provide bi-directional train protection?
• Should a "standard" be established for CBTC train position resolution, to support operational requirements (or perhaps a range)? What about minimum train separation (i.e. distance to which a following train can safely close-up to a stationary train ahead)?
• Should a "standard" be established for commanded speed limit resolution, to meet operational requirements (e.g. 1 mph, less, more)?
• Can a "standard" functional approach be established for implementing temporary speed limits (e.g. speed limit resolution; definition of speed limit location)? Is it acceptable for temporary speed limits to be input/removed via a (non-vital) ATS system, or should other "standard" mechanisms be specified? Similarly, is it acceptable for track/switch blocking to be applied/removed via a (non-vital) ATS system?
• Can a "standard" functional approach be developed for definition of work zones (protection of track forces)?
• Can a "standard" functional approach be developed for moving trains into/out of CBTC territory?
• Should a CBTC "standard" also include interlockings, or should interlockings be considered independent of CBTC?
• Can "standard" functional requirements be defined for CBTC system check-out after maintenance and/or prior to entering revenue service?
• Can any functional "standards" be developed with respect to interfaces between the carborne CBTC equipment and other car subsystems, such as propulsion, brakes, doors, etc. (ref. WG#5)?
• If track circuits are retained for broken rail detection, functionally how should CBTC respond to/operate through an unexpected "occupied" track circuit?
• Can "standard" functional requirements be defined for cutting/combining trains?
• Can "standard" performance requirements be defined for CBTC automatic speed regulation algorithms (e.g. acceptable tolerances around "ideal" speed/distance profiles; acceptable variations around "ideal" trip times; etc.)?
• Can a "standard" stopping accuracy be defined for CBTC programmed station stop algorithms?
• Can a "standard" functional approach be developed for automatic train regulation (e.g. dwell time modifications; run time modifications; etc.)
• Can any "standard" functional requirements be developed for other CBTC-ATS functions (such as system status displays, automatic vehicle identification, train tracking, automatic routing control, definition of operating schedules, junction management, energy management, train hold functions, train stop functions, system performance monitoring, passenger information system interfaces, etc.)?
• Can any "standard" functional requirements be established in response to system failures (e.g. NYCT functional requirement to continue to provide civil speed enforcement in the event of loss of RF data communications with the wayside)?

These issues were briefly discussed by the Peer Group. At the conclusion of the discussions, the Peer Group were then asked the following questions:

• Do you see value in developing standard performance/ functional specification for CBTC systems?
• Do you believe it is practical to develop such a standard specification?
• Would you actively support IEEE Working Group #2 in developing such a standard specification?
• Would you actively promote the use of a consensus standard specification in future CBTC procurements?

The Peer Group responded unanimously in the affirmative to each of the questions. The Peer Group commented that while developing a standard functional/performance specification is practical, the effort required to reach consensus on the many operational issues should not be under estimated. The Peer Group also requested continued TRB support to provide financial assistance to cover travel costs to attend the working group meetings.

Tom McGean advised that the Peer Group’s support of the proposal to develop a standard performance/functional specification for CBTC systems would be reported to the IEEE standards committee at their meeting in Washington on October 8, 1997. Tom McGean also advised that if the committee endorsed the proposed new direction for Working Group #2, then Alan Rumsey has offered to take over the chairmanship of the Working Group for this endeavor.

The meeting concluded with a tour of PATH’s Operations Control Center.

All of the meeting participants expressed their thanks to Pete Furkey and other representatives of PATH for their hospitality in hosting the Peer Group meeting.

It was agreed that the next meeting of the CBTC Peer Group should be scheduled for the day after the next Working Group #2 meeting. The following dates were proposed:

Wednesday, November 5, 1997: Working Group #2

Thursday, November 6, 1997: CBTC Peer Group

It was proposed to hold the meetings in the New York area, possibly at the Long Island Rail Road. Pete Furkey also offered to hold the meetings at PATH, if required.

[Post meeting note: Victor Grappone has confirmed that LIRR would be prepared to host both the Working Group #2 meeting and the CBTC Peer Group meeting. Note, however, that the date for the Working Group #2 meeting has been changed to Friday, November 7, 1997].