My interim CMM colleagues have not been using this blog, but that doesn’t mean nothing has happened. It has been a long while since last blog entry and for now I’ll focus upon present status.
We have been “slipped a pole” on delivery of annual CIM information model releases. A combination of factors caused this and it may have been unrealistic to expect we could achieve nice clean annual delivery of standards. There are interoperability tests (IOP’s) and real world projects that tend to drive things more than any standards body discussions or high minded goals of delivery cycles. I definitely think it is the right decision to be more goal focused instead of just doing annual releases, but it complicates our nice clean CMM role rotation. Anyway, I have assumed the UML model management role and coordinating the weekly WG13 issues meetings since late January and am beavering away at cleaning up remaining issues in the “CIM15” release.
The “CIM15” or more precisely the IEC61970cim15v15 UML package was “frozen” in the January WG13 meeting. Here “frozen” means we won’t do things that break existing profiles unless there is a very serious issue. This was really started at the October 2010 combined WG13+WG14 meeting where there was backlash on changing a few names. So we are doing a bit better job at not breaking implementations. The CIM15 IEC 61970-301 doc will go out later this year.
At the same time we are fixing critical problems (mostly omissions) and are up to iec61970cim15v21 as of today. Most of the changes have been in the “Dynamics” package that came from EPRI work and has been used by ENSO-E in preliminary IOP’s last year. We also added some new primitive types like “Date”, “DateTime”, “Time”, “Duration”, and “Decimal”. These are needed to more precisely map to XSD based messages and to properly represent financial data without round off errors. We have been addressing issues for the upcoming ENTSO-E IOP (July) and for the WG14 IOP in late March.
A big trend coming now is a move from focus on information model to focus on profiles. This has been a somewhat challenging transition and will continue over the next year at least. Part of the “pole slip” is due to this where in 2009 and 2010 years the focus was much more on profiles than on advancing the information model. We are still learning how to best manage this and there are not any really good examples or textbooks we can follow.
There may also be a more formal recognition that CIM is used for standard messages as well as for enterprise integration scenarios as a Canonical Data Model (CDM). This seems to be a trend and usage that may have been under appreciated in the standards bodies.
We expect an IOP of “CIM15” this summer but there are no firm plans and it is getting fairly late. There is planned to be a new 61970-452 CPSM (CIM power system model) profile this spring to support the ENTSO-E IOP if nothing else. Some minor changes to 61970-456 (solution exchange) profiles will occur as well.
The most dramatic development in the information model last year has been the introduction of new transformer models which are not completely new or different, but a significant change. These changes enable combined transmission distribution models, balanced or unbalanced representation and ability to use the same model (both instance and meta data) to represent both types of systems. That may not be clear, but it’s the key enabler to avoid every backend having to keep mappings of how transmission models link into (or overlap with) distribution models. It also allows unbalanced analysis on the transmission system to be layered on instead of having to simply use a completely new more detailed model. Much of the credit on the new transformer model goes to WG14.
The IEC61968 has had a huge amount of work in preparation for the upcoming March IOP’s and the maturity of those models is greatly increased. A lot of things were made “informative” or simply removed to make way for this maturity.
The IEC62325 model has also matured to support EU markets and this year the North American models are poised to also move forward with much the same process used by WG13 and WG14.
There is a full day CIM panel session at IEEE PSCE in Phoenix week of March 21.
I’ll say more on “CIM16” plans in a future blog entry, but for now CIM16 is on hold until we finish preparations for the upcoming IOP’s. Be sure to use the CIM issues if you have questions or identify problems.
Just a quick summary of CMM perspective in 2008.
We made good progress on the Planning integration. This also includes progress on solve state exchanges and case input.
Lots of other ideas were discussed in the issues meetings but movement is slow, but the basis for many 2009 activities has been built.
The UCTE project in latter part of 2008 and early 2009 will likely be a pivitol development and we have done a good job supporting that activity so far.
Quite a few items that have been lingering problems have surfaced and the good news is they are being addressed.
We have more issues than ever, but we also knocked out quite a few. The outstanding issues are a good focus including many directional issues as well as details. Its clear that more people are examining CIM with a critical eye now and that is very good.
The 2008 "inter op" was a success. The concept of ModelingAuthoritySet exchanges was proven to be workable at a basic tool level. The model changes introduced for IOP have survived to the CIM 13 draft standard.
CIM 13 (IEC61970CIM13) is basically frozen unless the Baden WG13 meeting opens some serious issue. The 61970-301 document is in draft state and available on this sharepoint to CIMug members under the IEC draft documents.
Our transition to Sparx Enterprise Architect tool was completed. WG13 and WG14 were working closely together to good effect.
I did not focus much on WG16 in 2008 but I think that will change in 2009, as all parties are more ready for that integration.
We have a good list of 2009 activities to be prioritized this week in Baden by WG13. This week also includes WG16 meeting in Seattle (Kirkland) which will further shape the upcoming work for CIM related TC57 WG's.
We had a first real customer issue submitted on the SharePoint model issues list. Hopefully that will get traction as hearing directly from users is needed. See the button ont he main CIMug page.
The transition of CMM role to Kurt Hunter is taking place now and I am confident in his abilities and previous experience in the CMM role. My best wishes to Kurt.
Many thanks to those who have supported and endured my tenure as CMM. I look forward to contributing to the CIM community, but hopefully with quite a bit less travel.
2008 CIM Model Manager
We have been making changes the last few weeks on the "Frozen" 13r07 IOP model. The changes will not impact IOP in October as all changes are outside the network/powerflow model exchange profile (IEC61970-452 aka CPSM). I believe the IOP group could take the latest draft models (up to 13r11 this week) without any changes to the exchange profiles.
Some more changes related to 61968 package changes and merging shared components into 61970 have also occured. Gradually the 61970 diagrams are also shaping up a bit, but there is still more work to do on these. I will be focusing on IOP model preparations in the next few weeks and taking some vacation.
It has been hard to get a concensus group on our weekly issues calls because of summer vacations, but we have made some progress on having WG14 and WG13 coordinate. Tatjana Kostic has been great in assisting and organising the WG14 issues for discussion by WG13.
We keep ticking off a few issues each week and if we can keep that up, we may stop the growing backlog of open issues. I feel we are making progress, and improving the situtation, though we clearly have lots of work ahead.
Complex data types, implications on UML model and tooling will be a major topic. Also the discussions on operational limits and state exchange are continuing, though we have the basics of those modeled already. The existing operational limits UML model should be sufficient for IOP this year.
If any users or vendors out there are already building on top of the TopologicalNode class, please let us know as this might get reworked a little to better support state exchange use cases.
The challenges with the new CIMug site and the associated IEC sites has been a litttle frustrating, but when it works its great! Hopefully people will hang in there and keep contibuting to the sites.
I will plan to do some more work in the next few months on the tools WG site, but if anyone wants to add content there regarding a particular tool, please go for it! Go to main CIMug page, select "Groups" button near top of page, then select menu item "CIM Tools Site". Then click the "Tools Wiki" button on the left side of page. Here you could go to "CIM Tools List" page hyperlinked from the main content. Here you can do your Wiki style edits to add content on a tool by selecting the "Edit" button on the upper right of page.
We have a set of changes to "CIM13" (more precisely IEC61970cim13r07) which will be used for the 61970 "inter op" testing in October 2008. The IOP group has agreed to "freeze" the model to be used for this testing. By freeze, we mean no changes to the part of the model that is in the IEC61970-452 exchange profiles (aka CPSM - CIM Power System Model). Also we will of course unfreeze and correct anything that is a serious problem for IOP.
In the meanwhile, the IEC TC57 WG13 (and WG14 and WG16) will continue to evolve the annual CIM release, but the IEC will not lightly make modifications that impact the IOP "frozen" model. There are number of model cleanup issues, like documenation, and diagram cleanup that don't impact the IOP. These areas will continue to evolve throughout 2008. Also there are areas of the model (for example state exchange) that are not part of the IOP profile(s) and will continue to evolve in 2008 and likely be part of the official IEC61970-301 standard targeted for the beginning of 2009.
We are targeting to do some limited testing of the ModelingAuthoritySet this year as well as the usual model exchanges. We are also targeting testing exchange of Equivalents package which was introduced last year in CIM12.
Work continues on the dynamics models within the EPRI sponsered CIM for planning - dynamcs activity. There is not yet final agreement on UML changes, but the process appears to be converging on a solution that enables exchange of dynamic models in much the same way as we exchange static models. In fact, the dynamics models will be an additional modeling layer that associates back with the static model much like a various packages like Siemens-PTI PSS/E and the GE PSLF have handled their dynamic and static models in their proprietary file forms. DigSilent (PowerFactory) has also been contributing heavily to this effort.
WG14 is working closely with WG13 now and intends to do next edits with the combined EA model.
WG16 European CIM Market Extensions are yet to be merged but I hope to get to that soon, though its unclear what exactly we will be doing with the contextual model aspects that have been performed in UML.
Well we are managing the model now in the Sparx Systems EA tool. We are successfully merging iec61968 models from Rose, though it is not trivial to do so. Things will get much better once we are totally within the EA tool. The only outstanding item is the line routing on the diagrams, but we will be doing a final merge and manual cleanup this year.
The iec61970 models are being completelyupdated within the EA tool. Some manual cleanup of the diagrams within 61970 has already started, but this is a fairly low priority and can be done later in the fall after we InterOp.
There is a new Rose plug-in available to export an XMI file that is much better for EA to import. The latest versions (7.1) of EA are also needed for maximum compatibility with the new Rose XMI export. With this combination and a little extra processing by a tool called CIMinEA.exe y we can perform a valid clean import.
We have a potential problem with matching guid’s used in the XMI files not matching, but the CIMinEA.exe tool and another called merge_guids.exe can be used to match things up by package and class names instead of by guid.
I will post some more on this CIMinEA.exe tool later as we release these tools in open source form.
Slides on this process and tooling were presented at the IEC working group meetings in Vasteras.
The meetings the last two weeks in Vasteras Sweden went very well. We had a good turnout for the European CIM users group and many presentations and some good discussion during the “ask the experts”. One of the major achievements was in the CIM Naming and Model Exchange working group breakout, where an action plan was created for the UCTE. The CIM community considers the UCTE a exchanges a key driver and we had presentations from UCTE and members.
There appears to be general agreement that the process, tools, and ability, to derive contextual models and messages from the information model is key and is now a central issue in making good progress.
I was able to attend one day of the WG16 meeting and we will be working this week (after returning to the office) on merging the European CIM Market Extensions into the “CIM13” model. I should say combined 61970cim13_61968cim10_combined model. This work merge work will of course be done in the Enterprise Architect tool. I will talk more on EA migration in a separate post.
The combined WG13 and WG14 meetings were driven mainly by WG13 issues meeting topics, and the WG14 attendance was a little smaller than usual. The later joint meetings suffered from WG14 members breaking into smaller working groups to do the things they needed to accomplish. However there was good progress and generally agreement. I think we are making good progress working closely with WG13 and WG14 jointly participating in weekly issues calls for the last several months. We are also managing the WG14 and WG13 issues in the same IEC CIM issues list. That IEC issues list will be maintained separately from the CIM ug issues list, but the CIM ug issues list will feed into the IEC list.
The WG13 meeting was a big one to resolve outstanding items that are needed for the InterOp tests this October. A number of issues were resolved, closed, and some really good collaboration and UML modeling were accomplished. The changes and agreements were specific enough for me to put out an IEC internal 61970cim13r04 release to be reviewed by the WG 13,14,16 teams this week. I am hopeful that this UML model is very close to being frozen for IOP 2008 in October.
We did agree that some items outside the IOP 2008 scope may evolve during the latter half of the year. These changes will not impact the parts of the model used for IOP.
It is looking good for CIMTool to create the profile artifacts (OWL or RDFS) that the vendors will use for IOP 2008.
IOP discussions this morning are debating details of ModelingAuthoritySet testing. Though it looks like there is interest from some vendors. A few of us have been pushing the ModelingAuthoritySet testing since it seems critical to UCTE and other key drivers in the industry to utilize CIM exchanges.
We have over a hundred model change entries now. Though this is not final and WG13 could nix some of these yet, the following are key items for “cim13”.
- Aggregated regulation control and regulation cleanup.
- New equipment operational ratings package.
- New load type model and voltage response capabilities.
- New state variables package.
- New control area specification package
- Planning style bus (TopologicalNode) labels.
- Branch group interface specification
A combined team from WG13,14,16 will be working on model repackaging also. Since packaging does not impact IOP, this can be done in the latter part of the year.
Also, we hope to get to some major documentation cleanup, and fill in some undocumented or weakly documented parts of the model. Again this does not impact IOP and can be done in the fall.
We are migrating from old site to new, but there is a problem migrating blog entries. So I am going to post some old entries here today.
Title: Recent WG13 meeting and CMM role summary
Our WG13 CIM issues phone discussions will start to take place in about a week. This is where much of the real decision making on CIM changes and additions takes place. The recent WG13 meeting included presentations and some discussion of the “CIM for planning” work sponsored by EPRI and the ERCOT CIM extensions. Both include lots of good work with some issues to be resolved on how this would move into the CIM standard for more general consumption.
Presently I am advocating within the WG13 and as CMM to get out of managing and distributing our model in a weakly supported, difficult to install, and extremely expensive UML tool. The majority opinion appears to be in favor of moving to a tool called Enterprise Architect from SparX as a capable but much less expensive alternative. The recent WG16 and WG13 meeting showed general approval of such a move, though it must be done cautiously and with concern for:
1) Producing the 61968-301 document
2) Keeping good diagrams which have been a pain point in transferring from one tool to another
3) Retaining the ability to merge models produced by various groups into the “combined” model.
4) Supporting existing Rational Rose users who have committed to building tooling on top of that platform
Your inputs on this issue are welcome.
Another topic that came up recently is releasing documents in PDF format. I have submitted an issue on this and will be raised in the IEC working groups.
Back to the title….
I am the CIM Model Manager for 2008, this is a role officially within the CIM Users Group, but recognized by the IEC as having similar parallel responsibilities. The role is by agreement to be supplied on a rotating annual basis (roughly along the lines of annual CIM releases) by each of the three major interested vendors (Siemens, ABB, AREVA). The vendors are cycling in the order specified and 2008 is AREVA’s turn. If you disagree with the selection of major vendors, then you can request to join in the rotation. A CMM candidate is supposed to at least participate in most of the major CIM UG and IEC WG13 meetings over the course of the year prior becoming CMM. The CMM is automatically (or at least normally) included as a member of WG13, 14, and 16.
Specifically, the CMM role has the responsibility of facilitating gathering CIM issues, facilitating discussion of the issues both in official IEC meetings and with series of phone conferences, documenting and implementing the resolution of these issues into the draft CIM UML model. The CMM does not get to officially decide anything, but as implementer of changes is often free to choose concrete implementations for review by the TC57 IEC groups. It has been my observation that good ideas are generally well received in the standards groups, though might take some work to clearly communicate. The CMM also merges the various models together into a “combined” UML model and generates the annual 61970-301 document. CMM also answers specific questions raised in the context of the CIM Users Group help desk. The CMM gives reports of progress to both the IEC working groups and the CIM UG.
Title: EA import from Rose
I am continuing to work with the IEC groups toward our CIM13 goal of maintaining the model outside Rose and likely in Enterprise Architiect. We are making progress on that but we discovered (maybe you knew already) that the Rose to Enterprise Architect imports are loosing all attribute references to classes like “ActivePower” in the Domain package. The most visible indication of this problem is the extra “Domain” classes that appear at the root package level and the “<undefined>” class might also appear there. In fact all attribute classifier references were being lost.
This is due to a bug or non-conformance to XMI standards in the Rose export. The latest Rose may have fixed this (I don’t know), but the Rose copies most people have are not likely to be upgraded and have the bug. Fortunately we have great automation tools in EA and I wrote a script that takes advantage of the fact that we don’t duplicate class names in our combined CIM model. Thus the text attribute types are translated into proper references. This issue is hard to see from the GUI UML tools since they typically only show the attribute type name and it looks the same whether it is just text or a reference.
You can see the error by choosing the attribute classifier selection tool in most tools and it will show the reference class highlighted. If no class is highlighted, likely you just have text. This appears to be the functionality in RSA, StarUML and EA. I have to check with the lawyers before I can release the EA patch code or program, but I’ll ask if I can do it to help everyone out.
Now that I have cracked the EA automation, I may write some more validation tools to help us keep the model more consistent. I have posted the patched 12r02 release in EA project file on the IEC WG sharepoints for review and will make sure it gets to the CIMug site when that standard is released.
We will pick up on the organization and explanation of our processes at the start of this new year. I think it is fair to say we have not yet got all the kinks worked out. We want to make it easy to get information, review and support issue resolutions, and create a new issue. We are also working out what files we are legally allowed to present on this sharepoint. Stay tuned as I expect activity will pick up at the beginning of the year on streamlining the web site.
Side note: the IEC WG 13 is conducting a “paper interop” on display exchange standards after the January meeting (paper IO on Jan 21-22), so anyone with use cases for display exchange should submit those to Erich Wuergler who is spearheading that event. See the public document at http://www.uisol.com/cim/WG13/Documents/Common_Graphics_Display/CommonGraphicsExchange-NWIP-R2.doc for a contact and some background.
Side note2: The IEC WG 13 will be looking closely at the “CIM for Planning” proposed additions at the January 16-18 meeting also.