.15926 Roadmap

November 25, 2010

Roadmap of .15926 (ISO 15926 implementation project, see https://levenchuk.com/2010/11/14/dot15926-iso-15926-implementation-project/):

Version 0.5 — full fledged template engine

* Full support of Part 2+ language, Part 7 templates.
* N-Tree browsing and editing (with templates lowered and lifted)
* drag-and-drop template creation and editing
* no interface that pattern-defined by user
* smart auto-completion in all fields
* search for pattern with result in any branch of a tree (with auto-completion on base of previous searched patterns)
* Server and Editor work on one computer
* No constraints language.
* No DSL support at editors.
* Adapter to OWL files only (rudimentary mapping) import/export (Part 8 )
* Intraserver (not from RDL Federation) ontolet support
* No project/preset support
* No server optimizations
* smart installer

Version 1 — diagram editor + RDL server
* Part 7 diagrams browsing and editing (lifted and lowered graph supported)
* Constraints language support
* Multiple adapters (full mapping support)
* Remote server
* Participation in RDL federation (SPARQL endpoint publishing, Part 9)
* Presets/project support
* No collaboration (simultaneously editing) support

Version 1.5 — collaborative ontology editing environment
* collaborative editing

Version 2 — universal modeling environment
* scalable server (multiple optimizations)
* multiple DSLs support (language workbench: projectile editor)

After study of several (Proteus, iRING, Bentley etc.) implementations of ISO 15926 we at TechInvestLab started new ISO15926 implementation project that called .15926 (think about MS .NET multilanguage technology). Why we performing “yet another” implementation? As usual, because we wanted to program/model in ISO 15926 “language”, but there was (and still is) no suitable IDE for programming in this “ontology-based programming language”.

There exist rather large (now it have 46 subscribers) Russian-language community of this project at http://www.livejournal.com/users/dot15926/). This post introduce .15926 to English-speaking people.

There is several subprojects:

.15926L — ISO 15926 “language” of Part 2 and Part 7 (language is also called .15926L). .15926L consists from Part 7 template language and logical “constraint language” (like in UML+OCL pair).This pair of sublanguages are supported by “virtual machine” for storage and processing of .15926L data, implemented by publish-subscribe paradigm server. Server is one of member of RDL federation and can be running in local computer or somewhere in the Net. Internal presentation of ISO 15926 data is “ISO15926 native” (not triple-store) but there will be multiple adapters, one of it is to OWL/SPARQL to be fully compliant with parts 8 and 9. Other adapters exist for mapping to foreign schemes. Server support language paradigm of “pattern matching”.

.15926N — different notations for “ontolets” (ontology modules for different experts domains). These notations and ontolets are define DSL (domain-specific-language). All these DSL (including Part 2 ontolet and diagramming notation, Part 7 ontolet and template implementation-independent notation) are supported by IDE (interactive development environment) that is of “language workbench” kind.

.15926O — ontolets set that is “ready to use out from the box”. These are support implementation-specific data (e.g. “ontolet”), several upper level ontolets standards for system definition (ISO15288), situational method engineering (ISO24744), architectural descriptions (ISO42010). Also .15926O have “refactored” classes from these standards.

.15926M — methodology for work with ISO15926 that is using description standards from .15926O ontolets (i.e. situational method engineering and architectural descriptions standards).

There will be several versions of .15926LMNO components, that eventually reach to declared in this post descriptions. First versions will not have many of planned features (e.g. server will not support constraint language, IDE will not be a language workbench etc.). Programming language for first version of server+IDE is Python. Actual coding started since 1 of October 2010. After reaching “beta” state .15926 software will be released under open source license, .15926O will be submitted to RDS/WIP (JORD).

This is slides for my keynote talk “Approach to Enabling System of Systems Engineering” that I plan to give tomorrow (18th of October 2010) in Moscow at International Workshop on System of Systems at ICUMT2010 (http://icumt.org/images/programma_verska_2.pdf).

RuSEC 2010 presentations

September 27, 2010

Presentations (both English and Russian versions of each one) you can find here: http://www.vniiaes.ru/files/RuSEC%202010.rar

RuSEC 2010 results

September 26, 2010

At EuSEC 2010 in Stockholm (May 2010) INCOSE European chapter leaders stated that «INCOSE SE VISION 2020 should be updated by bottom-up procedure starting from local chapter’s visions of SE challenges». This work was already started by France and Russian chapters.

INCOSE Russian chapter in 2009-2010 have had 32 biweekly regular meetings and one 2-day Workshop (March 2010) to discuss SE challenges in Russian language. Systems Engineering Challenges Workshop (RuSEC 2010, http://rise-russia.org/rusec2010) with international participation was 23-24 of September’10 in Moscow, Russia.

INCOSE Member Board co-chair Asmus Pandikow was with us in Moscow and delivered a talk in behalf of INCOSE Central and Europe.

In average there was about 70 people attended most complex talks of second day. Workshop started with 350 engineers and engineering managers, but after first day most of the managers and engineers leave due to very high technical level of discussions – we were not aimed for pop and easy SE but hard core one, that correspond complex nature of SE project in 2020..

RuSEC 2010 had successfully brought together leaders of three communities: systems engineers, situational method engineers, industrial ontologists. At this Workshop INCOSE Russian chapter proposed a project that can help to capture knowledge about SE 2020. We need to:

  • Provide structured description of SE (Body of Knowledge) as a cohesive set of method components that can be assembled for a specific SE project to form a particular SE method (example of this form of representation you can find at http://opfro.org that have more than 1100 systems engineering method components. Chairman of OPEN Consortium that delivered this SE knowledge capturing and structuring project Donald Firesmith was present at RuSEC 2010 and agree to participate with initial content of this SE method catalog).
  • Provide that these method components fit to ISO 24744 development process metamodel. Developers of ISO 24744 were present at our Workshop and agree with our initiative.
  • Provide that these method components expressed in form of ontology to participate in federation of semantic Reference Data Libraries (industrial ontology-based knowledge bases that defined in ISO 15926). Project manager of devoted to building operational top-level Reference Data Library JORD project of POSC Caesar Associations and FIATECH consortium (https://www.posccaesar.org/wiki/PCA/Filatech) and developers of ISO 15926 was present at our Workshop and agree to collaborate with INCOSE in establishing this SE method RDL.
  • Provide that content of this systems engineering knowledge base correspond not to SE 2010 (like BKCASE and SE Handbook) but to updated SE VISION 2020 and have answers to systems engineering challenges. This is primary goal of RuSEC Workshop: have a model-based and ontology-based vision of model-based and ontology-based systems engineering (happy to have this «meta» in description).

This project can combine ontology form and SE content with situational method engineering as foundational glue between the form and the content. All presentations delivered at RuSEC 2010 tell us that this project is extremely hard to complete but there is no fundamental obstacles to this.

Current INCOSE SE VISION 2020 INCOSE-TP-2004-004-02, V2.03, September 2007 (http://incose.org/ProductsPubs/pdf/SEVision2020_20071003_v2_03.pdf) need to be updated to reflect advances in SE that was in place since 2007 when this vision was approved.

INCOSE Russian chapter suggests (http://www.slideshare.net/ailev/systems-engineering-challenges) that systems engineering body of knowledge should be expressed according to ISO 24744 metamodel not only in case of already established knowledge but also in case of challenges. According to ISO 24744 development method metamodel we can structure discussion of SE challenges in following focus areas:

Meta-level (SE as a whole):

  • SE vs. “Plain Engineering”, Art, Science.
  • Systems thinking (notions of System, System-of-Systems, holon, etc.)
  • SE Pragmatic (mitigating risks only vs. positive value adding).
  • Let’s borrow ideas from software engineering immediately, not after 10-15 years delay.
  • «Brave new SE» (a blend of model-based, semantically-enabled, cost-based, generative, SoSetc.) vs. “traditional SE”.
  • Demystification of “SE Art”: SE Knowledge discovery vs. SE Knowledge design (Engineering of Systems Engineering).
  • SE Knowledge modeling/capture and management: Situational method engineering (OMG SPEM, ISO 24744, SEMAT) and SE Method catalogs (method elements repositories – OPFRO) and BoK (BKCASE).
  • SE Ontology (System-of-Systems debate, 4D-debate, etc.).
  • SE Vocabulary (national/industry specific).
  • Industry-specific adoptions of SE. «Defense accent» of SE.

StageKinds (TimeCycleKinds)

  • TimeCycle (LifeCycle) management methods: Agile vs. Waterfall; Lean, Six Sigma and TOC in SE; etc.
  • TimeCycles coordination for SoS, multi-contractor work (e.g. DEMO organization engineering framework).
  • Project Management in SE (inside or outside)?
  • What school of project management (PMBoK, PRINCE2, P2M, TOC, LastPlanner, constraints planning, etc.)?
  • TimeCycle methods with Integral Teams, Virtual Collaboration, etc.

WorkUnitKinds(SE disciplines, practices, processesetc.)

  • Traditional set for modern (model-based, semantically-enabled, cost-based, generative, etc.) SE :
  • requirements engineering
  • safety and security engineering
  • Engineering of system architecture
  • Design
  • Configuration management
  • Integration
  • Verification and Validation
  • …..
  • Or completely new set of disciplines?!
  • TBD (e.g. High-level modeling)
  • TBD (e.g. Low-level modeling)
  • TBD (Generative manufacturing)

WorkProductKinds:

  • Megamodel as first class WorkProductKind (configuration management of megamodel)
  • System of Systems / Family of Systems
  • Technology Platform/Generation and technology as system-of-interest
  • Human-based system (including enterprise engineering)

ModelKinds, Languages (Metamodels), Notations:

  • Multiparadigm modeling (SysML, ModelicaML, OPDM, knowledge representation languages)
  • Engineering/Architecture Metamodels/Frameworks (UPDM etc.)
  • Engineering DSLs (like P&ID) vs. Universal Engineering Languages (like SysML or AADL)
  • Semantic/ontology-based megamodel integration (Simantics, ISO 15926, etc.)
  • Specific metamodels for SE disciplines:
  • Requirements discovery/engineering (e.g. stakeholder models, GORE frameworks and standards, etc.)
  • Cost models
  • Engineering of system architecture (UPDM, MFESA etc.)
  • Generative design
  • Assurance case, design rationale and other reasoning capture models

ProducerKinds

  • People
  • Roles of systems engineer (requirements engineer, systems architect, configuration manager, V&V authority, security and safety engineer etc.). How to build SE team with these roles?
  • SE Education (continuous education: 1 year experience repeated 20 times should not count as “20 years of experience”)
  • Tools
  • CAD/CAM/CAME/PLM/Wiki/Social networks/etc.: what services/tools will support SE in 2020? Data-centric vs. Document-centric, Semantic vs. Vocabulary, Model transformation vs. Model editing, etc.

[This post was submitted to be published in INCOSE INSIGHT]

Speakers from UK, USA, Spain and Finland:

23 september 2010, Moscow, Russia

My draft slides for talk at RuSEC 2010 (http://rise-russia.org/rusec2010):

Moscow today

April 27, 2010

image

Test post from HTC Desire.

The order in this language list is important (from programming to modelling then to ontology-based modelling):
1. SIMULA67, BETA and Scandinavian OO-modeling school (programming=modeling)
2. Modelica — object-oriented acausal equation-based (multy-physics) simulations — http://modelica.org
3. ModelicaML — simulations in Modelica and systems modeling with SysML (http://www.openmodelica.org/index.php/developer/tools/134). Less integrated but with the same intention is CATIA V6 RLFP+BDM (with Modelica and SMC) from Dassault Systemes.
4. SysML and other languages for systems modeling.
5. OPML — ontological product modeling language (http://www.nist.gov/cgi-bin//get_pdf.cgi?pub_id=904119) and SysMO initiative (http://www.sysmo.org, now seems not too active).
6. OIM (object information model) and their integration in ISO 15926 (http://www.infowebml.ws/intro/introduction.htm).

I realize that for each line in this list there is a certain narrow community. Each one of such communities is rather well familiar and has discussions with neighbours but no dialogue exists between distant members of this list.

ISO 24744 (with addition of ISO 42010) can be used as a framework for discussion of all of these languages and notations of modeling in engineering and development activities. Using these standards with connection to modeling language spectrum list should permit to build MBSE process framework that would explicitly provide explanation for usage of different types of modeling in different tasks and processes of model-driven development project life cycle.

I am a president of INCOSE Russian Chapter.

INCOSE Russian chapter have biweekly meetings, there was 22 such a meetings since Chapter creation in a spring 2009. All the material (including video of presentations) is available here: http://community.livejournal.com/incose_ru/ (all meetings are in Russian).

INCOSE Russian chapter have now 44 members from Moscow, Sankt-Petersburg, Ekaterinburg, Kiev. We use http://dimdim.com and http://skype.com to gather via internet our members from different cities.

Occasionally we have presentation from our 23th of December 2010 meeting translated in English — http://www.slideshare.net/ailev/situational-method-engineering. Regard it as a sample of an INCOSE Russian Chapter meeting presentations.

Main topics of interest in our chapter is:
— international standards of systems engineering and situational method engineering
— model-based systems engineering
— requirements engineering
— project planning for capital construction projects
— CAD/PLM
— configuration management on base of ontology data integration
— magister curriculum in systems engineering and other education activities

There was 2 international tutorials in Rosatom-VNIIAES (“Systems thinking and systems engineering” and “Data integration”) that was attended by Russian chapter members (materials in Russian see at http://www.vniiaes.ru/uzhc/).

We plan to provide Workshop on systems engineering challenges (25-28 of March 2010, near Moscow, Russian language). Special topic will be planning activities in capital construction projects (we have 4 speakers for this type of challenge).

23-25 of September 2010 we plan to organize international conference on challenges of systems engineering (RuSEC2010).

My main goal as a president of INCOSE Russian Chapter is to provide presentations of Russian engineering work at international level. Out members should to publish articles about their understandings in systems engineering in English, they should to attend international events as speakers etc.

Two times in ontologies

January 8, 2010

LinkedIn have “ISO 15926” group, and this is my words in discussion “Accepting and using class definitions in non-15926 terms”:

Current computer/modeling revolution tend to be in the “interactive programming/modeling-in-large” (see http://www.slideshare.net/kruder396/evolution-in-the-large-andnd-in-the-small, http://en.wikipedia.org/wiki/Interactive_programming etc.) direction that is the same thing that difference between object (ontology) and method (epistemology) discussion here. From Method domain classes are NOT time-independent (there is Schema versions in all CAD suits, and painfully conversion of information models from old Schema to new one). From Object domain classes are time-independent, but this is different time.

I think that future version of ISO 15926 (ISO 15926+) should be reflective ( http://en.wikipedia.org/wiki/Reflection_%28computer_science%29 ), to include Method discussion, not only Object discussion.

There is two orthogonal “times” — first is of “class model development” (that is invisible from time in 4D) and second “in 4D”. I consider that we need conceptual framework to explicitly discuss both these times, “epistemological” and “ontological”. 4D (and ISO 15926) is conceptual framework only for “ontological/object” time, but modelers need “epistemological/method” time (that should be added to known and shared ontology and labeled appropriately) — at least for ontology configuration management…