Proposal for the Subdivision of Project 7.19
for a Petri Net Standard

Contribution to ISO/IEC JTC1/SC7/WG11 SEDDR

Date: 14 August 1995
From: Australia
Subject: Subdivision of Project 7.19 for a Petri net standard
Work Item: 7.19 Diagrams for Software Engineering

Project 7.19.xx Petri Net Standard

Preface:

The development of an international Petri net Standard has been discussed in ISO/IEC JTC1/SC7/WG11 at its meetings in Ottawa (June 94), Bonn (November 94) and Brisbane (June 95). At its meeting in Brisbane, WG11 agreed to the subdivision of Project 7.19 as the appropriate project for the proposed standard. WG11 submitted a subdivision proposal to the SC7 Plenary, but withdrew it as it was felt that the proposal should be more widely circulated.

The purpose of this contribution is to seek National member body comment on this proposal, in order to proceed with a subdivision.

1. Title:

Petri net Techniques

2. Scope and Field of Application:

The scope of this subproject is the definition and standardization of a Petri net technique including graphical notation, syntax and semantics. Notions of high-level nets and augmentation with time for performance evaluation are within the scope of this project as are techniques for the analysis of Petri nets and the development of a transfer syntax.

The technique can be applied to a wide variety of concurrent systems, in particular distributed systems. The fields of application include

3. Purpose and Justification:

3.1 Background

Petri nets are a technique for the specification and description of the dynamic behaviour of systems and in particular those systems that express the concurrent operation of various subparts. The technique has an intuitively appealing graphical form, which facilitates understanding of the behaviour of systems by humans.

Since their invention in the early 1960's, substantial development of the techniques for the modelling, specification, analysis and testing of a broad range of systems has occurred over the last 15 years. This has included extensions for performance analysis (Stochastic Petri nets, Timed Petri nets) and for concise representations (High-level nets, which include abstract data).

Nets have been applied to a broad range of systems, particularly those that may be implemented in software, including operating systems, databases, communication protocols, manufacturing systems, defence command and control, business processes and telecommunications. They have been used in both government and business organisations in a large number of countries, on six continents. Examples include: specifying telecommunications services (Deutsche Telekom, Germany); modelling industrial systems (HP, USA); Multimedia services (LAAS, France); Application to Railway traffic control (Netherlands); modelling communication protocols (Telecom Australia); Communication software development (Oki Electric, Japan); Petri net tool development (Denmark); Flexible Manufacturing Systems (Spain, France); Large scale document storage (CIT,Germany); Object orientation (Italy); railway signalling (Australia); Claims Processing (Meta Software, USA); Production Line Control (Kobe Steel, Japan); Protocol testing (GMD, Germany); space communications (ESA, France); safety critical systems (PARSE project, UK); computer performance modelling (Hitachi, Japan); business modelling (DMR, CRIM, Canada); dependability and reliability (USA); protocol performance (South Africa, France); computer supported co-operative work (Italy); distributed systems (Finland); and banking (USA) to name a few.

There is an annual Petri net Conference held at the end of June (the next in June 1996 will be held in Osaka, Japan) and a biennial workshop on Petri nets and Performance Models, which started in 1985. There is a Petri net mailing list on the internet (PetriNet-request at daimi dot aau dot dk) and also a world-wide web page (http://www.daimi.aau.dk/PetriNets/).

In order to develop industrial applications, tool sets for Petri nets are essential and include: graphical creation and editing; simulation; functional analysis; performance analysis; testing and automatic implementation (compilers). There are now a number of commercial tools available (eg Eval from Verilog, France; Design/CPN from Meta Software, USA; and PROMPT from Unico Computer Systems, Australia), as well as a large list of tools from research organisations. A list of these tools is kept on line by the Canadian organisation CRIM, which can be accessed via the world wide web.

3.2 Specific Aims

During their development, many different variants of Petri nets and their extensions have occurred. The purpose of the standard will be to provide a reference definition that can be used both within and between organisations, to ensure a common understanding of the technique, and the specifications written using the technique. This will allow the standard to be referenced in contracts, facilitating international enterprize and trade. It will also facilitate the development and interoperability of Petri net computer support tools.

3.3 Beneficiaries

The main beneficiaries will be in many industrial sectors, including telecommunications, defence, aerospace, banking, computing and transportation. Petri nets can also be used in understanding business processes and in their re-engineering and this will effect many secondary and tertiary industries including government service organisations.

A Petri net standard would facilitate the exchange of specifications and designs written in Petri nets across national boundaries, thus assisting trade.

It would also facilitate the development of compatible tool sets.

3.4 Feasibility

The development of the standard is technically feasible. The techniques are quite mature with several text books available in a number of different languages.

3.5 Timeliness

There has now been over 30 years of development of the techniques from research ideas to a level of maturity necessary for industry uptake. This has been happening over the last 5 years. In order to facilitate: understanding of Petri net specifications and designs between suppliers and clients; and the development of compatible tools, the time is now right to produce a Petri net standard that is useable by industry.

3.6 Urgency

Numbers of Petri net tools are being produced, most of which are incompatible. In order to facilitate the electronic exchange of designs and specifications in Petri net form, it is important that the standard is produced quickly. The schedule for the standard is given in the Programme of Work.

3.7 Benefits

The benefits of a Petri net standard are many including:

  1. The use of the standard in contracts between purchasers and suppliers, and between suppliers and their subcontractors. This will allow unambiguous specifications to be made which are independent of the language and culture of different countries.
  2. Use of the standard will allow much better understanding of the behaviour of the desired systems, by allowing visualisation of both flow of control and flow of data within a system.
  3. It will allow specifications and designs in Petri nets to be transferred between Petri net tools designed and developed by different organisations.
  4. It will allow the development of standard analysis techniques, to ensure certain properties of systems are captured by the specifications.
  5. It will allow standard Petri net building blocks to be developed for re-use.
  6. It will facilitate debugging of specifications and designs at the earliest opportunity, before implementation, thus saving costs, and increasing quality of products and services.
  7. Facilitating tool support for rapid specification prototyping.
  8. Facilitation of a market for Petri net tools.

4. Program of Work:

4.1 Status

A draft outline of a Working Draft for the standard was produced at the last SC7/WG11 meeting in Brisbane in June 95 (ISO/IEC SC7/WG11 N182). This draft is currently being filled out by work in both Australia and Germany. It will be circulated with this document.

4.2 Schedule

The following schedule is proposed.

Project		  Working Draft	  Committee Draft	DIS	IS
----------------------------------------------------------------------
7.19.xx	  	  96-05		  97-05			98-05	98-12
----------------------------------------------------------------------

4.3 Resources

The following resources have already been committed in 4 countries.

Country		Commitment
---------------------------------------------------------
Australia	Editor
Denmark		Work on Transfer syntax
Germany		General contribution to Petri net standard
USA		General contribution to Petri net standard
---------------------------------------------------------
Other resources are expected from many other countries for reviewing the standard, once the activity is advertised more widely.

5. Reference Documents:

  1. J. L. Peterson, ``Petri Net Theory and the Modeling of Systems", Prentice-Hall, N.J., 1981.
  2. W. Reisig, ``Petri Nets, An Introduction", EATCS, Monographs on Theoretical Computer Science, W.Brauer, G. Rozenberg, A. Salomaa (Eds.), Springer Verlag, Berlin, 1985
  3. K. Jensen, ``Coloured Petri Nets'', Volume 1: Basic Concepts, Springer-Verlag 1992.
  4. K. Jensen, ``Coloured Petri Nets'', Volume 2: Analysis Methods, Springer-Verlag 1994.
  5. J. Desel and J. Esparza, ``Free Choice Petri Nets'', Cambridge Tracts in Theoretical Computer Science 40, Cambridge University Press, 1995
  6. R. David and H. Alla, ``Petri nets and Grafcet'', Prentice Hall, 1992.
  7. A. A. Desrochers and R.Y. Al'Jaar, ``Applications of Petri nets in Manufacturing Systems: Modelling, Control and Performance Analysis'', IEEE Press 1995.
  8. Advanced Course on Petri Nets, Bad Honnef, West Germany, September 1986. Published in `Advances' series, LNCS Vols 254, 255, 1987.
  9. E. Best and C. Fernandez, ``Notations and Terminology on Petri Net Theory'', Arbeitspapiere der GMD 195, March 1987.
  10. J. Billington, `'Many-sorted High-level Nets", in K. Jensen, G. Rozenberg (Eds.), High-Level Petri Nets: Theory and Application, Springer-Verlag, 1991.

6. Cooperation and Liaison:

Co-operation with the Steering Committee for the International Conferences on the Application and Theory of Petri Nets. Potential liaison with ISO/IEC JTC1/SC7/WG9 and IEC TC56 Dependability Aspects of Software.

7. Preparatory Work

7.1 Project Editor:

Australia offers to provide the project editor:

Professor Jonathan Billington
University of South Australia, Australia
Fax +61 8 8302 3873
ph: +61 8 8302 3940
email: <<j dot billington at unisa dot edu dot au>

7.2 Work on Standard.

Canada, Germany, Japan, USA and Australia have provided input to the Standard during WG11 meetings. Denmark has also committed resources to working on the standard (see section 4.3).