IEEE RAIL TRANSIT VEHICLE INTERFACE STANDARDS COMMITTEE

WORKING GROUP 2

Communications Based Trains Control Performance/Functional Requirements

Minutes of Meeting

Attendees:

1.0 MEETING OBJECTIVE

The objectives of this meeting of the IEEE Rail Transit Vehicle Interface Standards Committee (RTVISC) Working Group #2 (WG2) were to follow-up on action items from the previous WG2 meeting, held on March 25, 1998, and to commence a more detailed review of Sections 6.1 (Automatic Train Protection functions) and 6.2 (Automatic Train Operations functions) of the proposed standard, draft D3.0, issued on May 17, 1998.

2.0 HOUSEKEEPING ITEMS

2.1 Introductions

Alan Rumsey, WG2 chair, welcomed participants to this meeting of WG2, developing performance and functional requirements standards for Communications-Based Train Control (CBTC) systems. This meeting was hosted by MUNI, and Alan Rumsey thanked Patty DeVlieg and the MUNI staff for all of the meeting arrangements.

Patty Devlieg provided a brief summary of the status of the MUNI ATCS project. The MUNI Metro Turnback (MMT) and MUNI Metro Extension (MMX) Shuttle between Embarcadero and the CalTrain Depot have been in operation since January 1998, with AUTO mode ATCS between Embarcadero station and the portal (Stage 1). The remainder of the Alcatel SELTRAC ATCS system on the existing subway line is currently under final testing and will enter full revenue service shortly (Stage II). Through service between the existing subway and MMT/MMX (Stage III) will be cut-over later this summer (August 22) and the final software release will be implemented by end of this year.

Alan Rumsey noted that a meeting of the CBTC Peer Group was being held after the WG2 meeting, on July 16/17, also at MUNI, and all transit agency representatives were invited to attend.

Alan Rumsey also noted for the information of the meeting participants that the draft WG2 standard was currently being used by SEPTA as the basis for their procurement of a CBTC system for their light rail tunnel. Interest in the WG2 standard has also been received from as far a field as Australia (Westrail), and Europe (Czech Railways). Jacques Valerio (Matra) has also offered to promote the work of WG2 at meetings of the U.K. based Institute of Railway Signal Engineers (IRSE).

Alan Rumsey advised that he presented a status report on WG2 at the recent APTA Rapid Transit Conference in San Diego, and that he would be presenting a paper on North America initiatives on CBTC standardization at an upcoming conference in Lisbon, Portugal (Comprail ’98).

2.2 Working Group Membership

Alan Rumsey advised that the current WG2 mailing list consists of approximately 100 active members and friends, excluding certain individuals who had either asked to be removed from the mailing list, or who did not currently appear to be actively involved in the work of WG2. The current mailing list was tabled, reviewed and updated by the meeting participants. (The updated mailing list will be posted on the web site, www.tsd.org).

Of the approximately 100 active members and friends, 45% are users (transit agency representatives and their consultants), 35% are suppliers (including consultants to suppliers), with the remaining 20% being other interested parties.

2.3 Ballot Group

Alan Rumsey advised that the group’s target was to be in a position to ballot the CBTC performance and functional requirements standard by the end of this year and it was therefore appropriate at this time to consider the make-up of the ballot group. Two options are possible. The first option is for the full Rail Transit Vehicle Interface Standards Committee to act as the ballot group. The second option is to form a ballot group from the membership of WG2. Given the size and diverse make-up of the WG2 membership, and recognizing the specialist nature of the WG2 subject matter, this second option becomes feasible, subject to the approval of the RTVISC. One additional advantage of this latter approach is that WG2 would have more direct control over the schedule for balloting the proposed standard.

These options were discussed by the meeting participants. Marty Lukes (WMATA) proposed that the RTVISC be requested to permit WG2 to form a ballot group from within its membership, with interested members of the RTVISC who are not currently members of WG2 also invited to be included within the ballot group. This proposal was seconded by Bill Petit (Safetran) and unanimously adopted by the meeting participants. Alan Rumsey will carry forward the groups recommendation to the RTVISC.

2.4 IEEE Standards Association

Alan Rumsey advised that the IEEE has established a newly empowered operational entity - the IEEE Standards Association (IEEE-SA). The IEEE-SA Board of Governors has ruled that members of standards sponsor ballot groups formed as of June 1, 1998 must be IEEE-SA members to cast a ballot.

Application forms for IEEE-SA membership were made available for the meeting participants.

2.5 Date/Location of Next WG2 Meeting

George Anagnostopoulus offered to host the next WG2 meeting at the Volpe National Transportation Systems Center in Cambridge, Massachusetts and it was agreed to hold the meeting on Wednesday, September 30, 1998. The next WG4 meeting will be held at the same location on Thursday, October 1, 1998.

[Post-meeting note: The next CBTC Peer Group Meeting, for transit agency representatives only, will be held in Boston on Tuesday, September 29, 1998, hosted by the MBTA (John Lewis).]

2.6 RTVISC Meeting in Chicago

Alan Rumsey advised that the last meeting of the full Rail Transit Vehicle Interface Standards Committee was held in Chicago on June 17/18, 1998. The next meeting will be held in St. Louis the week of September 14, 1998, to coincide with the AAR meeting that week.

At the Chicago meeting, it was noted that Bill Petit (Safetran) and Bob Heggestad (Harmon) will serve as committee liaison to the FRA Rail Safety Advisory Committee (RSAC) regarding their Positive Train Control (PTC) efforts. Bill Petit advised that minutes of the RSAC meetings would be posted on the Safetran web site.

The RTVISC chair, Tom McGean, had also reiterated the committee policy on Definitions. Namely, it is committee policy to use TRB Glossary definitions if the IEEE Dictionary does not have a required transit term. However, TRB Glossary definitions are not consensus standards. Thus, the TRB definition should be provided in the definitions section of the standard so it can be balloted and adopted. A further policy of the committee is that once a definition is balloted and accepted for any standard sponsored by the RTVISC, it will be binding for all standards sponsored by the committee, unless it can be shown to be clearly inappropriate.

George Achakji (Transport Canada) had advised the committee of a report he has prepared for the Transportation Development Centre of Transport Canada, entitled "A Review of the State-of-the-Art Train Control Systems Technology. Copies of this report can be obtained from George, free of charge.

3.0 PREVIOUS WG2 MEETING HELD ON MARCH 25, 1998

3.1 Approval of Meeting Minutes

The minutes of the previous WG2 meeting, held in Washington, D.C. on March 25, 1998 were accepted as written.

3.2 Review of Meeting Action Items

Action items from the previous WG2 meeting were reviewed as follows:

3.2.1 Specific Values for CBTC Parameters

Alan Rumsey advised that Annex C of draft D3.0 of the proposed standard ("Typical CBTC parameters") reflects inputs received from Jeff Smith (Harmon). Additional comments on Annex C have been received from Bill Petit (Safetran), Jacques Valerio (Matra), Fred Childs (PATH) and Howard Gregson (DMJM) since draft D3.0 was issued, and these additional comments were tabled for the information of meeting participants. These additional inputs will be reflected in the next update to the draft standard, for review at the next WG2 meeting. Other suppliers were also requested to submit inputs to Annex C applicable to their CBTC systems.

3.2.2 Definitions

Alan Rumsey advised that Section 3.2 of draft D3.0 of the proposed standard ("Definitions") reflects all agreements reached at the last WG2 meeting.

No additional comments/suggestions were received by the WG2 chair regarding the definitions of "emergency braking" and "maximum service braking", as defined by working group #5 (WG5). Alan Rumsey advised that a number of definitions that had originated from WG5 have since been revised in response to ballot comments on the WG5 standard, and these revised definitions will be incorporated in the next draft of the WG2 standard. These revised definitions were tabled for the information of the WG2 participants.

A definition for "Movement Authority" had been suggested by Harvey Glickenstein and was included in draft D3.0 for review by WG2. There were comments from the group that perhaps this definition was too general and should be made more CBTC-specific.

The TRB definition for "headway" had also be added to draft D3.0.

Other comments noted with respect to the Definitions section of the proposed standard were:

1. for WG2 purposes, it was suggested that a note be added to the WG5 definition of automatic train supervision (ATS) to indicate that ATS functions may also include routing functions
2. it was recommended that the "note" added to the TRB definition of reliability be deleted
3. definitions may be required for terms such as "Safe Braking Distance", "Maximum Authorized Speed" and "Continuous Communications", for example; in general, however, it was agreed that the group should first focus its attention on the body of the standard, and then re-visit what additional definitions may be required for clarification purposes.

Section 3.2 will be updated in draft D4.0 to reflect the above comments/agreements.

3.2.3 Operating Modes

Alan Rumsey advised that Section 4.5 of draft D3.0 of the proposed standard ("Train operating modes") reflects all comments received at the last WG2 meeting.

With respect to the action item that transit agencies were to provide input as to their expectations regarding various levels of degraded modes, to include definition of operating modes currently employed on their operating property, only one response was received, from John Lewis (MBTA). John’s inputs were tabled for the information of the meeting participants.

Other transit agencies were encouraged to provide inputs regarding their expectations in this area.

3.2.4 Headway Requirements

Alan Rumsey advised that Section 5.1 of draft D3.0 of the proposed standard reflects all comments received at the last WG2 meeting.

3.2.5 System Safety Requirements

Alan Rumsey advised that Section 5.3 of draft D3.0 of the proposed standard ("System safety requirements") reflects all comments received at the last WG2 meeting.

Jim Hoelscher (GRS) had provided revised wording to ensure compatibility with the draft WG4 standard.

Vic Grappone (LIRR) had provided inputs regarding proposed quantitative safety criteria, based on LIRR experience and expectations with respect to CBTC. Vic’s inputs were tabled for the information of the meeting participants.

A number of additional comments have been received on Section 5.3 of draft D3.0 of the proposed standard. Where possible, these comments will be addressed in the next update of the proposed standard for further review by WG2.

3.2.6 System Availability Requirements

Alan Rumsey advised that Section 5.4 of draft D3.0 of the proposed standard ("System assurance requirements") reflects all comments received at the last WG2 meeting.

Other that the input provided by John Lewis (MBTA), per item 3.2.3 above, no other input had been received to date from transit agencies, regarding their expectations with respect to reliability/availability/maintainability of CBTC systems.

This topic was discussed further under item 5.0 below.

3.2.7 Draft D3.0 of Proposed Standard

Alan Rumsey advised that the remaining sections of draft D3.0 of the proposed standard had attempted to respond to all other comments received at the last WG2 meeting.

No addition inputs were received to date on section 6.1.15 ("Grade crossing warning"), and as such this action item will remain open. Vic Grappone advised that LIRR requirements were still being discussed.

4.0 REVIEW OF WG2 COMMENTS DATABASE

Alan Rumsey advised that all written comments received on draft versions of the proposed CBTC standard were being entered into the WG2 Comments Database (with the exception of minor typographical comments). A printout of the database was tabled and reviewed by the meeting participants. The printout identified all comments that had been open at the last meeting as well as new comments that had been added to the database since the last meeting. (Comments that had been closed prior to the last meeting have been archived, and were not included in the current printout. All comments are being retained, however, to provide an historical record of the workings of the group).

Since the database was printed, additional written comments had been received from Omar Rezzoug (Matra) and Bill Gallagher (Alcatel), which were tabled for the information of the meeting participants.

In summary, Alan Rumsey advised that approximately 250 written comments had been received to date of which 188 had been addressed and were considered closed.

5.0 REVIEW OF SECTION 5.4 ("SYSTEM AVAILABILITY")

In response to an action item from the previous WG2 meeting, George Anagnostopoulos (Volpe Center) tabled an updated discussion paper to include a delay value weighting factor in the proposed formula for determining CBTC system availability, which attempts to relate CBTC equipment reliability to the on-time performance of the transit system. While in principle the meeting participants saw value in defining a CBTC system availability performance measure that was directly related to on-time performance, there were however concerns regarding the practical difficulties of measuring the CBTC system contribution to passenger delays. George Anagnostopoulos offered to provide a worked example of how this may be possible.

Rob McHugh (SkyTrain) noted that Skytrain had originally used equipment reliability measures as an indication of system performance, but were now focusing more on passenger satisfaction measures. SkyTrain uses a delay threshold of 2 minutes (with 90 second headways), and attempts to relate each service delay to a specific subsystem.

Bob Miller (BART) noted that BART measure system performance by tracking on time performance on a daily and monthly basis. The delay threshold is 5 minutes at the end point of a given line. BART use weighting factors based on actual passenger counts on a given line or route. BART also attempt to track the source of each service delay.

Vic Grappone (LIRR) noted that LIRR use a delay threshold of 6 minutes, with minimum headways of the order of 2 to 3 minutes. Again, the practical difficulties of establishing and correctly accounting for the source of a service delay were noted.

A 5 minute delay threshold is used by Metro North and New Jersey Transit. PATH use a 3 minute threshold, and MUNI will use a 4 minute threshold, with separate contractual requirements for service delay greater than 4 minutes, and delays of less than 4 minutes.

From the meeting discussions it was clear that any measures of delay needs to be related to the operating headways as delays at shorter headways have a more significant impact. Also, the impact of a failure depends heavily on the response to the failure.

Meeting participants also noted the importance of maintaining historical trends, so that an availability "baseline" could be established before introducing a new system.

In general, the suppliers represented at the meeting indicated a preference for a performance measure related to CBTC system availability, rather than a performance measure related to transit system service delays, primarily because of the difficulty in establishing a relationship between service delays and CBTC equipment reliability It was also noted that the system availability requirement effectively drives the level of equipment redundancy that needs to be provided; the higher the availability number specified, the higher the level of redundancy required, and the higher the cost. Simply stated, the challenge, therefore, is defining a level of equipment redundancy that will ultimately achieve the desired level of on-time performance of the transit system.

Omar Rezzoug (Matra) noted that system availability includes not only equipment reliability, but also the mean-time-to-restore (which may or may not include a mean travel time, depending on whether or not redundant systems with automatic switchover are available), as well as the operational and maintenance procedures. Improvements to system availability require that all of these areas be addressed.

Lou Sanders (APTA) commented that the discussions on availability and on-time performance represented a much larger issue than just CBTC, and suggested that it may be appropriate for this topic to be addressed by the full RTVISC, to ensure a common approach within all of the standards being developed.

Alan Rumsey suggested that, within the time frame allocated to develop this first CBTC standard, if the group was unable to reach a consensus on a performance measure that addressed both on-time performance (which was important for the users) as well as CBTC system availability (which was important for the suppliers), then an alternative approach may be to simply focus on defining those characteristics of a CBTC system that contribute to high system availability and improved on-time performance, such as redundant system architectures, graceful degradation, improved diagnostics, etc.. Vic Grappone suggested that defining availability requirements for specific CBTC functions may also be an option (e.g. percentage of time train has a valid speed command).

Alan Rumsey undertook to take the inputs from the meeting participants and update Section 5.4 in draft D4.0 for further review by the group at the next meeting.

6.0 REVIEW OF SECTION 6.1 ("ATP FUNCTIONS")

Section 6.1 of draft D3.0 of the proposed standard was reviewed on a subsection-by-subsection basis, and the following comments were noted:

1. with respect to the third paragraph in section 6., the term "continuous (geographic and time) coverage" will be replaced by "continuous geographic coverage"; the term "short headway train operations" will be replaced by "the defined performance requirements", and a statement will be added to the effect that the RF datalink shall not cause any electromagnetic interference to existing systems
2. in the first paragraph of section 6.1.1.1, the phrase "and as such shall also verify train length" will be deleted.
3. the third paragraph of section 6.1.1.1 will be re-written to clarify intent (i.e. no "lost" trains), and to also address train entry into CBTC territory
4. with respect to section 6.1.2, a new section and/or diagram should be added to define/clarify a generic Safe Braking Model for CBTC, to include the relationship between the ATP profile, other profiles that may be used for ATO or for train operator speed commands, and the CBTC movement authority limit; a reference should be included at an appropriate place in the standard to train operator reaction times; the safe train separation function should also include the roll-back tolerance of the lead train
5. the last sentence of the last-but-one paragraph of section 6.1.2 can be deleted as this requirement is covered elsewhere
6. in section 6.1.3, item e) should be revised to read "the maximum speed that would enable the train to stop safely prior to the limit of the train’s movement authority or to slow down sufficiently to meet appropriate civil or temporary speed restrictions upon entering that section of track"
7. in the last two paragraphs of section 6.1.3, "penalty brake" should be replaced with "emergency brake"
8. the two paragraphs in section 6.1.6 should be combined
9. section 6.1.8 should be renamed "door opening control protection interlocks", and these functions should be at the option of the authority having jurisdiction; item a) should be split into two separate bullets; a note should be added to provide for selective door open enable for the case where the train length is longer than the platform
10. similarly, the departure interlocks in section 6.1.9 should be at the option of the authority having jurisdiction; the requirement should also apply to any stationary train, not just trains stopped at a station
11. section 6.1.10 should be re-written to clarify intent; specifically, when ATP profile is exceeded, CBTC will initiate an open loop emergency brake application; train subsystems will be responsible for bringing train to a stop within an assured stopping distance; it was the general consensus of the transit agency representatives that in the event that train speed is brought back within the ATP profile, CBTC should not necessarily invoke an irrevocable stop
12. section 6.1.11 should be re-written to avoid duplication of requirements, to separate control functions from status functions, and to define response to loss of switch status
13. the title of section 6.1.13 will be expanded to include work train protection as well as work zone protection; users were requested to define specific requirements
14. miscellaneous other minor typographical changes

Section 6.1 will be updated in draft D4.0 to incorporate the above agreements.

7.0 REVIEW OF SECTION 6.2 ("ATO FUNCTIONS")

There was insufficient time available to review this section of the proposed standard and as such this agenda item was deferred to a future WG2 meeting.

8.0 SUMMARY OF MEETING ACTION ITEMS AND OBJECTIVES FOR NEXT MEETING

8.1 Assignment of Actions

The following specific action items were recorded:

8.2 Objectives for Next Meeting

The primary objective of the next meeting will be to complete reviews of section 6.1 and 6.2 (ATP and ATO functions), and re-visit section 5.3 and 5.4 (safety and system assurance requirements), as well as Annex C.

9.0 ANY OTHER BUSINESS

Alan Rumsey then closed the meeting thanking the working group participants for their attendance and active involvement in the various discussions, and again thanking MUNI for hosting the meeting. Ed Mortlock and Booz, Allen & Hamilton were also thanked for providing an excellent lunch.

Minutes prepared by:

Dr. Alan F. Rumsey

Chair, WG2