PROPOSAL FOR COMMUNICATIONS PROTOCOL INTERFACE HARDWARE AND SOFTWARE DEVELOPMENT

Overview of Proposed Effort

Supported by the TCRPG-4 rail transit interface standards program, an IEEE working group has been active since 1996 developing IEEE P 1473, Communications Protocol Aboard Trains. The proposed approach is built around two existing protocols; EIA 709, a general purpose control network developed by Echelon under the trade name LonWorks, and IEC FDIS 61375, a train communication network (TCN) developed specifically for railways which addresses their need for deterministic, cyclic, time critical data and provides other engineering features unique to train operations.

Echelon's LonWorks is a U.S. developed protocol familiar to most subsystem suppliers with a wide user base, extensive programming tools, and economies attributable to widespread use. The IEC TCN protocol has the advantage of extensive engineering and testing to handle the specific needs of railway applications, which can save engineering and development costs involved with intercar communications. The combined approach is intended to provide the user with the ability to tailor the best system for his/her needs while still providing for interoperability. The transit community appears to be converging upon this IEEE approach which was successfully balloted this Fall with no negative votes. This standard will enable the transit industry to avoid the proliferation of protocols associated with the use of Local Area Networks for rail transit, avoiding a "Tower of Babel" where extensive costs are associated with communications between microprocessors in subsystems manufactured by different suppliers.

The key to the workability of the proposed IEEE approach is 1) that the two protocols be able to transparently intercommunicate with one another, and 2) that these two protocols be open; commercially available from multiple sources and available in source code format with a commercially available reference implementation. This will assure that the protocols can be implemented by anyone on any microprocessor platform. Intercommunication requires gateways between the EIA and IEC protocols. A gateway is needed which will enable vehicle subsystems using the EIA protocol to communicate with an IEC compliant train bus and which is freely available to the industry.

Parallel with the IEEE standards effort, the San Francisco Bay Area Rapid Transit District's (BART's) Advanced Automatic Train Control (AATC) program has deployed an Echelon based communication system between head and tail cars of AATC equipped trains. Early work indicates that noise is limiting achievable data channels and rates. As a result, BART has become interested in participating in an operational demonstration of an IEEE P 1473 Echelon/TCN application.

List of Partners

The Consortium for Communications Protocol Implementation (CCPI) for this project consists of the following members:

Organizations Providing Mass Transit

Bay Area Rapid Transit District

Businesses Offering Goods and Services to Mass Transit Operators

Bombardier

Echelon

Harmon Industries

WABCO

Adtranz

Union Switch & Signal

In addition, this project is endorsed by the Transit Standards Consortium. Letters of interest/commitment are enclosed with this proposal as Appendix A.

State of the Technology

The complete LonTalk protocol specification document, now referred to as EIA-709.1Control Network Protocol Specification, can be purchased as Part Number EIA-709.1 from Global Engineering Documents (GED) for $142 USD. The GED web site url is The Reference Implementation of the LonTalk protocol specification can be licensed from Echelon Corp. at no charge by signing the ANSI-approved license agreement. The ANSI-approved license agreement is available from Echelon Corp. With this information anyone with the appropriate technical skills can build a LonTalk compliant implementation and therefore a LonWorks compliant system. Key integrated circuits (Neuron Chips) are available from two sources and mature development tools, network management software and documentation are available off the shelf at published catalog prices. There are approximately 4000 developers of LonWorks products and systems and approximately 6 million LonWorks devices in real-time process control systems throughout the world including many transit and rail applications. No development effort is required in this area.

In the case of the IEC TCN protocol, the standard is internationally accepted and non proprietary. However, to date no reference implementation is commercially available. Siemens and Adtranz have proprietary implementations of the design and source code; other companies would have to undertake the development effort, and at the present time no single supplier in North America is willing to invest in the development. Therefore a project is required to develop a Reference Implementation of the IEC TCN protocol. This project will permit North American car builders to benefit from the many years of engineering, research and testing which have gone into developing the TCN standard which supports full interoperability. It will also open the world market, which is adopting the TCN standard, to not only North American car builders, but to North American based subsystem suppliers who are presently conversant with Echelon technology.

With regard to the gateway, a number of suppliers are involved in application specific ports between Echelon and TCN, but no commercially available general purpose gateway presently exists.

Work to be Performed

A project is proposed to 1) develop a reference implementation of the Train Control Network (TCN) specified by IEC 61375 FDIS and referenced by the IEEE1473 draft standard, and verify that the IEC 61375 train bus specifications are complete and unambiguous by executing a "Clean Room" reference implementation, 2) develop a gateway between the IEC 61375 trainbus and EIA 709 compliant subsystems, and 3) demonstrate Echelon/TCN/Echelon communications at BART using the gateway and TCN train bus. The resulting implementations would be made available to any supplier interested in developing IEEE1473 compliant products. This would promote the use of American rail subsystem equipment on TCN rail systems, both in North America and overseas and create healthy competition as well as assure interoperability among Local Area Networks on rail transit systems.

The proposed project will thus consist of three elements 1) Train Bus Development 2) Gateway Development and 3) BART demonstration.

Train Bus Development. This project element will develop the hardware and software required to build a train bus network prototype to IEC 61375 standards representing a four car train. The prototype will support all train bus functions including automatic train inauguration, bus redundancy and management services. Deliverables will include:

  • Train bus node hardware design. This includes schematics, bill of material and all information required to build the wire train bus node. The design will use commercial, off the shelf parts and components.
  • The train bus software stack. This is the train bus protocol software containing the source code and the complete software documentation.
  • Support tools. All basic tools necessary to configure the nodes, troubleshoot the network and validate the train bus functions will be provided.

Gateway Development. This project element will support development of a gateway to enable communication between EIA 709 compliant microprocessor subsystems and an IEC TCN compliant train bus.

BART Demonstration. This will be at least a four car train bus prototype containing the reference hardware and software and will be used to validate compliance with the IEC standard and communicate Echelon/based train diagnostics across the TCN trainbus via the gateway.

Physical and/or Operating Characteristics of the Innovation

This innovation will enable full implementation of the IEEE P1473 Communications Protocol with seamless interoperability between the TCN protocol adopted by the IEC and the LonWorks protocol being widely used in North America. It will capture the benefits of both protocols, a wide user base and low cost in the case of LonWorks, and a carefully engineered system for train communications in the case of TCN.

Development of Prototype Equipment, Process or Pilot Program

At least one four car train bus prototype will be developed containing the reference hardware and software for TCN and also demonstrating use of the Gateway for both LonWorks to TCN and TCN to LonWorks communications. An operational demonstration of digital train line communications using the IEEE P 1473 protocol in an Echelon/TCN/Echelon configuration will be conducted at BART.

Project Output

Project output will include the following:

  • Train bus node hardware design. This includes schematics, bill of material and all information required to build the wire train bus node. The design will use commercial, off the shelf parts and components.
  • The train bus software stack. This is the train bus protocol software containing the source code and the complete software documentation.
  • Gateway hardware design including schematics, bill of materials, and all information to build the gateway hardware.
  • Gateway software containing the source code and complete software documentation.
  • Support tools. All basic tools necessary to configure the nodes, troubleshoot the network and validate the train bus functions will be provided.
  • A demonstration prototype. This will be at least a four car train bus prototype containing the reference hardware and software and will be used to validate compliance with the IEC standard.

Project Schedule

It is estimated that this project will require 30 months to complete.

Total Project Cost

The project cost is summarized in Table 1. Total cost is estimated as $1,130,000. The $400,000 cost of the gateway development portion of the project will be provided by the supplier members of the consortium. The $250,000 for the Test Site and Support Personnel will be provided by BART and the supplier industry. Thus 58% of the cost will be provided by the consortium.

TABLE 1

TOTAL PROJECT COST

Train Bus Development

Train bus prototype hardware and material: $80,000

Preliminary analysis (architecture): 90,000

Detailed design 240,000

Test and validation 70,000

Total$480,000

Gateway Development$400,000

Test site and support personnel $250,000

Total Cost$1,130,000

Assessment Plan

The project will demonstrate successful two way communications over the TCN network and gateways. A test plan will be developed to validate that the implementation is TCN compliant and LonWorks compliant, and a separate test plan will verify the ability to communicate successfully through the gateway between LonWorks and TCN and between TCN and LonWorks protocols.

Relation to FTA Five Year Plan and Investment Principles

This program supports the following FTA strategic goals:

1)Safety and Security will be enhanced by eliminating the hazards associated with communication over a variety of nonstandard protocols and by enabling North American users to benefit from the extensive engineering and testing of intertrain applications using TCN.

2)Mobility and Accessibility will be enhanced by providing economical interoperable and standardized local area network systems for rail transit. This will enable the rail transit industry to capture the benefits from improved service and lower cost offered by the computer age.

3)Economic Growth and Trade will also be enhanced.This project will permit North American car builders to benefit from the many years of engineering, research and testing which have gone into developing the TCN standard. It will also open the world market, which is adopting the TCN standard, to not only North American car builders, but to North American based subsystem suppliers who are presently conversant with Echelon technology.

4)Human and Natural Environment will be enhanced by the reduced cost and greater affordability of rail transit, which reduces the need for automobile trips, saving energy and reducing air pollution in urban areas.

This proposed project fits the following FTA Five Year Plan Program Areas:

1)Equipment and Infrastructure: Rail Equipment and Systems, Communications-Based Train Control Systems (The protocol can be used for train to wayside communications for CBTC) , Design, Testing.

2)Fleet Operations: Transit Intelligent Transportation Systems

3)Planning and Policy: Intermodal Connectivity (The project will help integrate rail communications within the overall TCIP and NTCIP framework)

4)Professional Capacity Building: The project is a textbook case of technology sharing between European and United States R&D.

FTA Investment Principles. The project offers a short term high payoff benefit in that the protocol can be used in current railcar procurements to save money and increase reliability. Long term benefits include eliminating the potential for a Tower of Babel in rail communications.

Individual railway suppliers lack the resources to undertake this development by themselves. Furthermore, if they did so, they would have no incentive to make the results freely available to the industry. Thus a joint public/private research and development effort is warranted. The results of this effort will make the IEEE 1473 standard not just a piece of paper, but a complete protocol system. It has been estimated by the TCRP G-4 project that interoperable communications protocols can save the rail transit industry $70 million annually.

1