The minutes of the previous WG2 meeting, held at the Long Island Rail Road (LIRR) on November 7, 1997, were accepted as written.

Alan Rumsey advised that the IEEE Standards Board had requested wording changes to the proposed new PAR (P1474.1), to develop performance and functional requirements for communications-based train control (CBTC) systems. Following discussions with IEEE standards staff, the scope of the new PAR has therefore been revised as follows:

"This standard establishes a set of performance and functional requirements for enhancing performance, availability, operations and train protection using a communications based train control (CBTC) system".

The purpose of the new PAR has also been revised as follows:

"There are currently no independent standards defining the performance and functional requirements to be satisfied by communications based train control (CBTC) systems. A standard will enhance performance, availability, operations and train protection as well as facilitate new CBTC applications".

The revised scope and purpose will be reflected in the next update of the draft standard. Alan Rumsey thanked Tom McGean for his efforts in reaching agreements with IEEE standards staff on acceptable wording for the PAR.

Alan Rumsey advised that the following membership definitions have been adopted for WG2:

Participating Member: Regularly attends working group meetings and actively participates in development of the proposed CBTC standard; name included in draft/final standards when issued.

Contributing Member: Unable to regularly attend working group meetings but actively participates in development of the proposed CBTC standard through input of review comments, response to action items, etc.; name included in draft/final standards when issued.

Friend: Individual who is interested in the activities of the working group and who wishes to receive meeting notifications, meeting minutes, and other working group documentation, but who does not wish to actively participate in or contribute to the development of the proposed CBTC standard; may or may not attend working group meetings as observers; name would not be listed in the draft/final standards when issued.

The balloting committee has not been defined at this point, and will be defined by the full Rail Transit Vehicle Interface Standards Committee, unless this committee delegates the role to the working group.

WG2 members are also classified as "Users", "Suppliers", or "Other Interested Parties". Consultants are assigned to one of these three classifications depending on their primary source of income.

The current WG2 mailing list was updated based on the above definitions.

The next WG2 meeting will be hosted by SEPTA (John LaForce) and will be held in Philadelphia the week of April 13, 1998 (to coincide with the ASME/IEEE joint rail conference). Specific details regarding the exact date and place for the meeting are still to be confirmed. WG2 members and friends will be advised of meeting details, once finalized.

A CBTC Peer Group meeting (open to transit agency representatives only) is also planned the same week as the next WG2 meeting.

Tom McGean noted that the next full committee meeting of the IEEE Rail Transit Vehicle Interface Standards Committee (RTVISC) will be held in Irvine, California on February 4/5, 1998.

Alan Rumsey thanked those WG2 members who had submitted written comments on the initial "stawman" draft standard, D0.1 (Harvey Glickenstein, Dave Rutherford, John Swiecick and Howard Gregson), and on the latest draft standard, D1.0 (Bill Gallagher).

The latest draft standard has been updated to address most of the comments received to date. It was proposed that a mechanism be put in place to provide visibility to all WG2 members of all comments received and how these comments have been dispositioned. Alan Rumsey agreed to establish a database to record written comments received. (Note: to be manageable, the database will not be used to record all of the minor wording changes that may be recommended, as such changes will ultimately be reviewed by the full working group during the detailed line-by-line reviews. Database will instead record the more significant comments that need to be discussed by the full working group in order to establish a resolution).

Based on comments received to date, a number of general issues had arisen that warranted further discussion by the full working group. These issues are summarized below.

Alan Rumsey asked if the proposed CBTC standard should include design parameters, if the definition of such design parameters is considered beneficial by users/suppliers, and if there is a consensus on the specific design parameter. For example, some reviewer comments recommended that the draft standard be reviewed closely to make distinctions between prescriptive design requirements and functional/ performance requirements. Other review comments recommended that parameters such as required accuracy of train location, for example, be defined in the standard.

The consensus of the working group was that the focus of the standard must be on operating, performance and functional requirements. However, the standard is intended to be a tool to serve the industry and as such the working group should not be overly rigid in its interpretation as to what should, or should not, be included in the standard. There was seen to be some value in making users of the standard aware of key design parameters for CBTC systems, and in including typical values for these design parameters in an Annex to the standard. It was noted that the premature definition of certain design parameters could unnecessarily drive a specific CBTC architecture, and that it is more important to focus on functional requirements. Typical ranges of key design parameters could be included, as guidelines. It was also noted that since this standard was intended for a wide range of applications, with a wide range of train operating densities, the working group should be careful not to define requirements that may not suite a particular application. Also, the group should recognize that design parameters may change as CBTC technology develops over time.

A number of review comments have noted that interoperability remains an important functional requirement for many users, and hence should be addressed in the proposed standard. Other review comments have argued that since interoperability interface specifications are not part of the current WG2 mandate, this issue should not form part of the current standard. Alan Rumsey asked if the proposed standard should address functional requirements of the train-to-wayside communications subsystem, to establish a foundation for future interoperability interface design standards, or should the current standard avoid any discussion on interoperability and train-to-wayside communications.

A number of users noted that while interoperability between rail transit properties is not a specific requirement, interoperability is an important procurement issue within a transit agency to permit competitive bids. Users made comparisons to modern cab-signaling/ATO systems, as used at WMATA/MARTA/BART for example, where wayside and carborne equipment was originally procured as one package, but today wayside and carborne equipment could be procured separately because industry-accepted requirements for wayside-to-carborne equipment interfaces have been established. Users are looking for a similar degree of procurement flexibility with CBTC systems, and therefore the proposed CBTC standard should recognize that interoperability is an important functional requirement.

A number of working group participants noted that interoperability encompasses more that just wayside-to-carborne equipment interfaces. Section 6.4 of the draft standard will therefore be renamed "Interface Requirements", and this section will be used to capture consensus requirements related to interoperability, where they exist and where they are useful.

It has been agreed that ATS (and ATO) functions be treated as optional requirements, driven by the operational needs of the specific applications. One comment received on the draft standard was a recommendation that ATS requirements be excluded completely from the proposed standard. Alan Rumsey asked if the proposed standard should exclude all ATS functions, or should ATS functions continue to be included in order to highlight the ATS capabilities that can be supported by a CBTC system.

The consensus of the working group was that the proposed standard should continue to include ATS functions, particularly emphasizing those capabilities that can be provided and/or enhanced with a CBTC system.

Alan Rumsey asked if the proposed standard should address CBTC system test requirements, as had been suggested in one of the review comments.

The consensus of the working group was that while testing was an important consideration (both initial system acceptance tests, and testing after system modifications), it was premature to address this issue at this time, and that this issue should be deferred to a future standard. The current standard will, however, continue to address requirements related to CBTC diagnostic test facilities for equipment maintenance purposes.

Many of the review comments received to date related to the definition of terms used in the draft standard. A review of Definitions will therefore be included as an agenda item at a future WG2 meeting. It was specifically suggested that a definition for CBTC systems that had previously been agreed by WG2 be included in the draft standard.

In discussing issues relating to operating modes to be supported by CBTC systems, the working group reached consensus on the following issues:

While CBTC systems should have the capabilities to support fully automatic driverless train operations, issues associated with driverless vehicles are complex and encompass factors beyond the specific requirements of the train control system. It was therefore agreed that the current proposed standard should address only those performance and functional requirements related to attended automatic train operations, and any discussion on more stringent additional requirements necessary to support driverless operations be deferred.

It was agreed that the standard should address mixed mode operations as a normal operating mode i.e. operation of unequipped trains within CBTC territory. It may be appropriate for the standard to differentiate between:

a) mixed mode operations as a regular mode of operations

b) mixed mode operations during the transition period only, as a CBTC system is cut-in

c) mixed mode operations as a result of carborne equipment failures

It was agreed that there were limitations to the extent to which failure mode operations could be defined in the standard as specific failure management strategies will be dependent to some degree on the specific CBTC system architecture. In general, however, it was agreed that failure mode operations could be classified as:

a) those failures which affect all train operating within a particular zone

b) those failures which affect a particular train operating within any zone

There was general agreement that since there is an operational requirement to continue to move trains in the event of equipment failures (albeit at a reduced level of performance, and possibly at a reduced level of automatic train protection), the standard should address "fail operational" requirements as well as "fail safe" requirements. As a minimum, CBTC systems must provide equivalent levels of operational flexibility and operational safety during equipment failures as can be provided with conventional train control systems. A number of users also commented that they would be looking to CBTC to provide additional benefits, and more graceful degradation, without requiring complex and expensive fall-back systems.

There was considerable discussion on the "Auxiliary Wayside Protection Mode", as defined in the current draft standard, and the potential need for onboard CBTC equipment to continue to perform certain ATP functions (such as civil speed enforcement) in the absence of communications from the wayside. This particular requirement could possibly drive the CBTC system architecture as it would likely require onboard determination of train position, and onboard knowledge of the track profile ahead, at least for the zone within which the train is currently operating. Alan Rumsey explained that the rationale for this particular operating mode, which had been established as an specific operating need by New York City Transit, was to maintain continued train operations for as long as possible in the event of short term or longer term equipment and/or data communication failures, while continuing to provide full train protection either through the CBTC system, or through an auxiliary wayside system ("back-up" signal system), or through a combination of both systems, and with minimum reliance, if any, on adherence to operating procedures.

The group highlighted the importance of standardizing on operating terms and definitions. It was also suggested that a matrix or other graphical representation by be appropriate to show the relationships between the various operating modes.

Alan Rumsey agreed to update the draft standard (Section 4) to reflect the above discussions, for further review at the next WG2 meeting.

The working group agreed that establishing the minimum achievable headway for a particular line and a particular operating fleet involved many factors that were outside the scope of the train control system. (Alan Rumsey made reference to TCRP Report 13 which addressed this subject in great deal). The group therefore agreed that the proposed standard should only address the CBTC contributing factors to achievable headways, of which the most significant were:

a) location (both accuracy of measured train location, and resolution of movement authority limits for a given train)

b) speed (both accuracy of speed measurement, and resolution of speed limits established for a given train at a given location)

c) communications delays (nominal and worst-case transmission times of command/status messages between train and wayside and vice versa)

d) CBTC equipment reaction times (both for wayside and carborne equipment, and for various operating modes, as applicable)

e) CBTC system performance limitations (i.e. maximum number of trains that can be processed by the CBTC system, within a given zone)

Based on discussions under item 2.1.1 above, it was agreed that while the standard should highlight that these CBTC parameters have an impact on minimum achievable headways, it may not be appropriate to assign specific values to these parameters in the standard. It may however be useful to include typical values in an Annex to the standard. To this end, suppliers were actioned to define specific values for the parameters listed above, with respect to their CBTC systems. Denny Pascoe also undertook to develop a sensitivity analysis which would attempt to relate variations in these parameters to variations in minimum achievable headways.

Alan Rumsey will update the draft standard (Section 5.1) based on the above discussions and further inputs received.

Section 5.3 of the draft standard has been updated based on inputs provided by Dave Rutherford in response to an action item from the previous WG2 meeting. Dave Rutherford explained that Working Group #4 (WG4) was attempting to establish a safety verification standard for processor-based systems, based on the specific safety goal that had been established for that particular processor-based system. It was therefore recommended that the WG2 standard should be limited to a statement of the overall safety goal for CBTC systems.

A suggested safety goal is defined in the current draft of the CBTC standard, and is based on the concept of a "unit" of CBTC equipment that is encountered by a train traveling at maximum authorized speed for 1 hour. (Jim Hoelscher advised that a similar definition had been used by Dr. Michele Berieau, of the EEIG ERTMS Users Group, Brussels, at the World Congress on Railway Research in Florence, Italy, in November 1997.)

There was some discussion within the working group as to the adequacy of the proposed definition for the safety goal, in particular whether or not the definition adequately covered the number of trains operating on the line at the specified operating headway. There was also some discussion as to whether or not the standard should attempt to define the top level "hazardous events" that would be covered by the safety goal, and whether or not there was a need to clarify the definition of an "unsafe state" (e.g. given that an extended stoppage on the line could be considered an "unsafe state" if passengers were to disembark from the train). It was also noted that in defining a safety goal for CBTC systems, it may also be appropriate to recognize that the capabilities of a CBTC system are seen to offer possible enhancements to the overall safety level of a transit system.

Meeting participants were requested to provide specific comments on the proposed CBTC safety goal directly to Dave Rutherford (with copy to the WG2 chair). Dave Rutherford will update the draft standard (Section 5.3) based on the above discussions and further inputs received.

George Anagnostopoulos tabled a discussion paper on CBTC system availability requirements. There was insufficient time available to review this paper, however, and meeting participants were therefore invited to submit any comments directly to George Anagnostopoulos (with a copy to the WG2 chair).

Alan Rumsey encouraged all WG2 members to submit any addition comments and suggestions on the current draft standard (D1.0), in writing, prior to February 28, 1998. The draft standard will be updated to reflect comments received, and will be distributed to WG2 members and friends prior to the next WG2 meeting.

For the interest of WG2 participants, Vinnie Ferrando (GE-Harris) gave an informative presentation on the GE-Harris organization, highlighting their involvement in Positive Train Separation (PTS) projects and advanced computer-aided dispatching systems.

The following specific action items were recorded:

The objective for the next meeting will be to attempt to finalize key performance requirements associated with:

a) CBTC operating modes (ref. Section 4)

b) CBTC contributions to minimum achievable headway (ref. Section 5.1)

c) CBTC safety goals (ref. Section 5.3)

d) CBTC availability criteria (ref. Section 5.4)

Time permitting, it is also proposed to review the Definitions section of the draft standard (Section 3.2)

Tom Sullivan advised that in order to comply with IEEE requirements, a security password may be required in the future to access working group documents on www.tsd.org. Working group members and friends requiring access to this information will be given details of any security password requirements. Tom Sullivan also offered to explore the possibility of setting up discussion groups on www.tsd.org, to facilitate ongoing discussions between WG2 members on standards related issues, in between the WG2 meetings. (It was however recognized that not all WG2 members have internet access).

Alan Rumsey then closed the meeting at 5 pm, thanking the working group participants for their attendance and active involvement in the various discussions, and again thanking GE-Harris for hosting the meeting and providing an excellent lunch and refreshments.

Dr. Alan F. Rumsey

Chair, WG2