[Sea-sa] Meeting reminder - SEA-SA, RASDS subset, Monday, 8 Nov 21

Fri Nov 5 00:04:39 UTC 2021

Dear SEA-SA RASDS enthusiasts,

We have a working meeting scheduled for this coming Monday, 8 Nov.  The attached presentation includes the latest set of updates that I have at hand.  I have been making a “consistency” pass over the materials, intended to bring the various views up to a common level of details and perspectives.  To that end there are several new pages added and some changes to others for consistency.  These changes deal with several, but not all, of the “Forward Plan” issues identified in the notes from the last meeting, which are appended here.

I draw your attention, in particular, to the following pages:  21-23, 27-28, 34-35, 39-42, 45-48, 52, 58-59, 69, and 90-101 (SysML annex).

I’d also like to draw your attention to the Operations Viewpoint, pgs 61 – 66.  In looking at these diagrams, and trying to create a one page “this is how to represent this” diagram (mocked up on pg 63), like pgs 20, 26, 34, etc I felt like we had not yet really come to closure on just what this viewpoint was.  Please see pg 62 for the questions I think we need answers for.

I think we still have similar questions re the Physical Viewpoint.  We have stated some concepts of how these components tie to the other physical components (Nodes & Links) in the Connectivity Viewpoint, but these ideas are not yet “crisp” in my mind.   I’d like the aspects of this view to be more concrete.  For one possible approach to that see below where I talk about openCAESAR and OML.

The other area that still feels “soggy” to me, i.e. “not crisp” is how we model the terminology / ontology.  In these updates I have adopted treating these ontological representations as info models (which is what they are).  Felt nicely recursive, see if you agree.  It does still mean that we create all of these ontologies “by hand”, which in this day and age seems rather retro.  We’ve tried some different tools that use OWL and other approaches, but none has, as yet “sparked”.  I have just loaded Eclipse, and I think Ramon has too.  There are a variety of modeling plug-ins that may give us the traction we need, and the platform is a) free, b) works on Mac and PC, and c) has a substantial international community.  It’s being used to good effect by the CSS SM WG for their Functional Resource Models (FRM).  Perhaps someone will take on the task of digging into this and providing some consistent ontology models.  There is also a SysML plugin for Eclipse.

I also learned of a different, but related, body of work called Ontological Modeling Language (OML).  There are TLA several variations on “OML” so search for the full name.  And there is an open source project called openCAESAR.  To quote from one of the leads:

OML is also “free”. It’s provided by the open source project “openCAESAR” (http://www.opencaesar.io/)  with a specification at http://www.opencaesar.io/oml/.

OML (Ontological Modeling Language) is a language that allows you to model ontologies. OML ontologies are of 4 types:

·         Vocabularies: defines your domain’s concepts, aspects, relations, properties, rules (inference rules) with open-world semantics

·         Vocabulary bundles: combines several inter-related ontologies and closes the world for them to permit “description logic” reasoning

·         Description: defines your systems as instances using terms from some vocabulures above

·         Description bundle: combines your descriptions in a dataset that can be reasoned on

OML is a higher level syntax around OWL2-DL (which supports DL reasoning) and SWRL (ability to extend DL semantics with more logical inference rules).

We can give you a demo if you’re interested.

Full disclosure: This has been developed at JPL as part of the Integrated Model Centric Engineering (IMCE) project, and OML and CAESAR and parts of the underpinnings for this project.  They are using this whole framework in three different ways:

  1.  Framework for formally modeling (and checking for consistency) ontologies
  2.  The IMCE modeling framework, which is an ontology
  3.  Using the IMCE framework and ontology for modeling mission systems and components, and checking those “realized” models for consistency with the ontologies  (variation of pg 5 in the attached)
On the one hand, this is a higher level of rigor than we are attempting to achieve, and they have a team of people working on this. On the other hand, one of the specific things they are modeling is all the spacecraft physical elements and creating a Mass Element List (MEL) and a Power Element List (PEL).  Anyone familiar with space missions knows that these are key gate products, and here they are being created from checked and validated models.

If anyone has an interest in looking more deeply into this and exploring the possibility of using this framework to assist our work I would support getting a demo and digging more deeply.  Don’t be shy, speak up.

See you all on Monday.

Cheers, Peter

RASDS++ working meeting on 18 October 21

Attendees: Huang, Krosley, Radulescu, Rovetto, Sanchez – Aguillar, Shames, Stangl, Vaden

Review of revised RASDS++ Powerpoint Materials (Function, Connectivity, Operation) – led by Shames & Radulescu

  *   Reviewed the 16 Oct 21 version of the RASDS++ PPT materials, with focus on overall coverage, consistency of presentation, terminology
  *   Stangl made the point that the ECSS has a separate working effort to define a set of Euro-centric terms in a Glossary.  He was requested to review those terms, and the RASDS proposed ones, and identify any overlaps or disconnects.  The point was also made that RASDS has chosen to align terminology to widely adopted and respected sources such as ISO 42010, NIST, TOGAF, DoDAF, and others.  The point was made in a later discussion (during the 20 Oct meeting) that in general there seldom is any single source that provides all of the desired terms and that some “detective work” and revising has been needed to create a single self-consistent, set of terms.
  *   Agreed during review to retain the viewpoint main object diagrams that show the main objects, input, output and management interfaces, main attributes, and concerns.
  *   During discussion of the Communications Viewpoint (protocols) agreed that we needed a simpler diagram in addition to the rather complicated one that was included as an example. This later led to adoption of this same general approach for all viewpoints.
  *   In the complicated protocol example that had been included we agreed (during 20 Oct discussion) to retain some of the security features, but to eliminate at least one of the three security approaches that were included.  All agreed that security is a very important topic, and that it had to be treated adequately.
  *   We spent some time on the Operations Viewpoint and reviewed the Activity Diagram that had been provided as an example.  This brought up a discussion of consistency in the use of object representations across different views.  This UML-derived Activity diagram adopted the UML style, which used rounded rectangles for Functions and rectangles for data.  But the RASDS unified “Legend”, pg 10, uses ovals to represent Functions and rounded rectangles to represent Data / Information objects.
  *   A separate working session was held by Radulsecu, Rovetto, and Shames, on 19 Oct, to explore ways to improve this Activity Diagram.  The result is in the 20 Oct PPT set, and it uses ovals, rounded rectangles, and dashed arrows, to align with the common practice documented in the Legend.
  *   Rovetto showed an ontology diagram produced with a tools called VUE and suggested that we explore use of it for the RASDS ontology diagrams.  Rovetto, Radulescu, and Shames explored this during their side meeting and concluded that this tool had some nice presentation features, but that it did not appear to have the ability to handle the OWL formulations that we have created.  It also only works on PCs and not on the more current Mac OS’s, so it is unlikely to meet our needs for an interoperable format / approach.
  *   => A common and agreed method for producing these ontologies is still desired, other options are solicited
  *   For the Operations Viewpoint we will also need a less complicated example diagram than the one provided
  *   The Activity Diagram, which is an example of use showing what is, in effect, a deep space style mission operations flow, was the subject of a lot of discussion.  There was a stated concern that it was not really accurate, that it did not include comm network activities nor science activities, and that it was missing some typical data flows.  All of these observations are likely accurate, but we have to remember that this is just an example of how this method may be used, it is not intended to accurately depict any given configuration or deployment.

SEA-SA RASDS working meeting on 20 October 21

Attendees: Huang, Krosley, Radulescu, Shames, Slane, Stangl, Vaden

Review of revised RASDS++ Powerpoint Materials (Enterprise, Connectivity, Operations, Service) – led by Shames & Slane

  *   Reviewed the 18 Oct 21 updated version of the RASDS++ PPT materials, with a continuing focus on overall coverage, consistency of presentation, terminology
  *   Reviewed the consistency in the package across different viewpoints, seeking to ensure a consistent presentation of different views and levels of detail
  *   Agreed on the following pattern for each viewpoint
     *   Each viewpoint will have a main object diagram that shows the primary object defined in the viewpoint, input, output and management interfaces, main attributes, and concerns.
     *   Each viewpoint will have a “drawing template” diagram showing two (or more) of the objects, the elements that they may include or directly reference, and how the usual relationships between them are depicted.  These are all in the abstract.
     *   Each viewpoint will have a constrained ontological diagram showing the main object, close “neighbor” objects, and closely related support objects in context, with relationships
     *   Each viewpoint will have a simple example diagram showing the use of the drawing template to create a relatively simple model
     *   Each viewpoint will have a more complex example view, probably drawn from one of the SCCS-ARD or ASL documents, showing use of the viewpoint methods to create a more complicated, realistic, view of some typical real elements, in context
  *   Motivated by the complicated protocol stack diagram, which includes a number of security protocols (data encryption, BP network layer security, and IPSEC IP layer security) we discussed removing the IPSEC, but also how to handle security in more general terms
     *   Agreed to do a pass across the whole document with “security eyes” on, to ensure that we have adequately supported cybersecurity modeling needs in the model methods & views that are provided
     *   Ensure that security topics were included in each complicated example where that made sense
     *   Keeping in mind that we are producing an architecture method and not an architecture, per se
  *   Reviewed the Enterprise Viewpoint with Fred Slane’s guidance since this is a specific SC14 concern
     *   Agreed to adopt many of the “enterprise architecture” terms that were provided, but to ensure that the term “Enterprise” was used in every case instead of “Business”
     *   Agreed that the term “Enterprise” was suitably broad enough to cover agencies, projects, multi-mission consortia, government organizations, businesses, and other commercial entities
     *   Reviewed the combines Enterprise / System ontology and agreed that this was a useful way of describing the relations ships (really correspondences) between objects in two different viewpoints
     *   This triggered a discussion of how these ontologies, classes, and relationships among classes relate to math, knowledge graphs, first order logic, and graph theory.  This is not a rabbit hole we will explore, but it does have relevance.
  *   Reviewed the Operations Viewpoint and Ontology
     *   Agreed that we need to provide the same sorts of intro diagrams for this VP as for the others in this set
     *   Agreed that the revised example provided by Rovetto, Radulsecu, and Shames aligns better with the rest of the RASDS++ style
     *   Agreed that we need a cleaner ontology diagram than either of the ones that we now have
  *   Reviewed the Services Viewpoint
     *   Agreed that the prototype service diagram was acceptable, but that it needed some clean-up (in draft form)
     *   Agreed that we should focus the Services viewpoint not just on SOA-style service paradigms, but also include options like on-board message bus, AMS, and web/cloud deployments
  *   Forward plan
     *   Add / update diagrams to meet agreed contents from this meeting
        *   Main object diagrams (all) – Shames
        *   Drawing template diagrams (all) – Shames
        *   Ontology models: Enterprise – Shames/Slane; Functional – Shames; Connectivity & Structural – Krosley/Slane; Operations – Radulescu; Information & Service - Shames
        *   Enterprise Examples – Slane
        *   Functional Examples – Shames
        *   Connectivity Examples – Shames/Krosley
        *   Structural Examples – Krosely
        *   Operations Examples – Radulsecu
        *   Information Examples – Krosley
        *   Service Examples – Shames/Slane
     *   Security “eyes” review – Shames/Radulescu/Sanchez-Aguillar

Next meeting – 2’nd Monday of the month, 8 Nov 2021

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/sea-sa/attachments/20211105/e05faf3d/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SEA SAWG RASDS Update Views 26Oct21.pptx
Type: application/vnd.openxmlformats-officedocument.presentationml.presentation
Size: 4773423 bytes
Desc: SEA SAWG RASDS Update Views 26Oct21.pptx
URL: <http://mailman.ccsds.org/pipermail/sea-sa/attachments/20211105/e05faf3d/attachment-0001.pptx>