2010-01-20 EA IP TSC WGM Minutes
TSC Enterprise Architecture Implementation Project (HL7 EA IP) Alpha Projects
Wednesday January 20, Q4 3:30-5 PM
- Hosting ArB, and Alpha Project teams for 2010JanWGM
- TSC / ArB opening remarks
- Call to order 3:37 PM by Charlie McCay
- Describe where we are from TSC perspective, anticipation going forward.
- TSC discussion Saturday revealed two activities:
- ArB creating framework,
- TSC to realize its value and make it effective within the organization.
- Recap of Agenda & Session Objectives - < 5 minutes; Marc Koehn
- Beginning the knowledge harvesting process
- We are now establishing the process. Communication needs to be improved.
- Sharing of information - We will be working on clarity of scope, efforts, and expectations
- Ron Parker notes it would be worthwhile to have dialogue at the end on the agenda.
- Woody has a prospective alpha project.
- Informing potential tasks for next cycle
- Marc notes the intent was to have some projects attempt to follow the approach described by the framework and feed back lessons learned into the process.
- Jobst asks if there will be opportunity to ask questions about the SAEAF itself.
- Beginning the knowledge harvesting process
- Alpha Project Updates (30 minutes)
- PASS (Privacy, Access, and Security Services) Don Jorgenson, project lead; Abdul-Malik Shakir, ArB Liaison
- Don Jorgenson begins with PASS update. First specification went to ballot; 50 affirmatives and one negative, reconciliation only took 15 minutes. Patrick Pyette lead author. Learned a lot along the way. When should be DAMs be DSTU? Surfaced questions on governance between projects and working groups, how voting and decisions are accomplished. Saw what the framework was offering us going forward; they see what they can do going forward to round it out.
- Marc asks for a one-pager from the group to note the issues that should be addressed by others.
- CTS2 (Common Terminology Services Release 2) - Russ Hamm, project lead; Cecil Lynch, ArB liaison
- Russ Hamm with CTS2 update notes they started a few years ago and kicked off prior to HSSP, and were a guinea pig to that process in SOA. Completed DSTU under the HSSP process and is now in OMG phase out for RFP. Looking at SAEAF to determine benefits can realize as an alpha as they move to normative phase. Looking forward to learning. Jobst says current SAEAF does not offer more than other tools like TOGAF and they’re looking forward to what the SAEAF development will offer. Proper SIM they have developed, as OMG has already adopted these frameworks.
- Andy Bond notes the SAEAF is not a replacement for existing EAFs but looks at the interoperability perspective the others do not offer, adding a layer to those frameworks. Marc notes we should be seeing the layering in DSTU and applying those layers to the evolution of those materials.
- Jobst will see PIMS from OMG by June 1st.
- Russ says during the interim CTS2 would have time to review the SAEAF perspective more properly.
- Infoway Blueprint 2015 - Ron Parker, project lead and ArB liaison
- Next major iteration of Infoway architecture using TOGAF and looking at the SAEAF stack and the SAEAF book at how they express and place artifacts in the stack; helps others understand what has to be done. Will stop short of actual interface design and message implementation. Traceability down to the service contract. Now finishing Bus req def’n; nothing to show you today. Blueprint SAEAF perspective will be visible here at HL7 before it’s visible publicly.
- Marc asks will we see SAEAF specific language within the context of the TOGAF framework that we’ll recognize? Ron notes htye will be learning that as they apply it. Even if we used TOGAF to build oru SAEAF we’d still have the architectural framework development to do. RM ODP doesn’t have it all, neither does TOGAF. Using it as a starting point for the expression of SAEAF does not preclude it.
- Don Jorgenson asks Ron if traceability and conformance ideas can be picked up in either framework; Ron thinks so.
- ITS- Project: The ITS
IsMight Be Broken - Dale Nelson, project lead and ArB liaison
- Dale Nelson on ITS says in September they tried to frame a new project based on “is the ITS broken”. Simplify it, make it harder, add things, remove things. Trying to capture requirements and come up with unifying ITS. Requirements gathering and governance bodies review have been underway. Formally started the project today, making sense of ECCF matrix and where ITS ‘plays’ in the matrix – they found they just couldn’t do it. Found approximate slots but couldn’t do it. Consumer of static expressions, expressing rules executed to build schemas and transforms for tooling to be developed against. They think they might be a behavioral component of the stack, expressing the behavior that occurs between slices of the layers. Another meta-level or dimension on the matrix and need SAEAF liaison help to clarify work products. It’s been an interesting start.
- EHR WG - Electronic Health Record System Functional Model Release 2 (EHR-S FM R2), Pat Van Dyke, project lead; John Koisch, ArB liaison
- Pat VanDyke talks about EHR WG, EHR-S FM R2, relaxed the ballot anticipation. Approach with development of use case to find where the pieces fit and run them through the stack.
- OO – Composite Order. Austin Kreisler, project lead; Patrick Loyd, ArB liaison
- Patrick Loyd with OO Behavioral Model and Behavioral Framework. DAM for a Behavioral Model, as compared to an informational model, writing storyboards whose interactions do not have payloads so are paradigm agnostic. Conceptualizing the Behavioral Model and BF and its pieces.
- Jobst asks about the modeling strategy, if it’s BPMN? Patrick says it’s Letter architect and Enterprise Architect for graphics, mostly on paper right now. Jobst notes for harmonization he’s concerned about the tool/representation.
- Structured Documents – Keith Boone
- Structured documents (Keith not here) but Patrick notes on the calls there was note that platform-specific and platform-independent complexities were identified so that clarification could be better achieved; discussions in the concept of SAEAF was a conversation that they had nothad previously.
- caEHR - Charlie Mead
- Charlie Mead reports on the NCI caEHR in Ambulatory Oncology. Will employ service specifications deployed incrementally.
- PASS (Privacy, Access, and Security Services) Don Jorgenson, project lead; Abdul-Malik Shakir, ArB Liaison
SAEAF specified services to be rolled out to 5000 ambulatory oncology centers.
- “Brainstorming”: Barriers/Challenges & Ideas – 45 minutes
- Marc says we’re in an experimental kitchen, adjusting the recipe as we go along. A number of projects enrolled are at different stages. How do we feed in the nuances that have been identified and roll it back out to the other projects? How to make improvements and changes to make it more applicable?
- Woody representing MnM notes the RIM as a mature product needs to find its place in the SAEAF model. Foundation group was asked to look at the things for which they are responsible to map them to the SAEAF. Would like to have the RIM augmented in time for May balloted to have it positioned in the enterprise architecture. Patrick says OO dynamic model expects changes in schemas and attributes. May not be ready to contribute those back to the RIM in time.
- What do we need to be thinking about as we move forward?
- Ann Wrightson notes in ITS they asked what constitutes a platform in the platform-independent vs platform-specific space? Have we ever really done platform-specific stuff that the MDA was intended to express? We’ve never really done platform-specific as the idea is interoperability between systems. It’s a bit like being metadata, different things to different people in different roles. Like the Java Virtual Machine that was never realized as a truly platform-independent tool.
- Charlie Mead agrees. Those terms came from OMG so they would not invent their own terms. SAEAF is trying to specify interoperability specifications. Conformance statements begin with a conformance assertion. Implementable is the key word, relativistic to one’s perspective. Conceptual, logical, implementable and the word ‘platform’ might be taking it too literally.
- Dale notes the other viewpoint on Engineering is on existing [platform capabilities, models, libraries etc. He asks where does the RIM, MIF etc exist in this matrix? If we look at this as a tool, is the ITS role to specify a computational viewpoint, moving one MDA row to another? How do we approach this, with the role of a transforming tool?
- Jobst notes implementability clarification is important. What is the set of instructions that can be generated from the exercise? Independence among the planes is not capable; some fields you cannot fill. State transitions of the MDA specs are a good reference.
- Charlie Mead notes we could spend the rest of the evening on this; it elucidates the big problem – how to specify the grammar you use to talk about it, then there is the implementation of a framework as an instance within an organization. We developed the framework and now have to develop our own implementation of the SAEAF. If we use the same grammar as another org that has an implementation of the SAEAF we can preserve the interoperability.
- Charlie McCay different audiences can speak architecture speak, or write specifications, and speak HL7-speak. ArB challenge must produce material that meets both types of speakers. How to engage effectively in those two communities and are these parallel activities or do we prioritize the dialogue?
- Marc reiterates there are arcane conversations at one level, and then others that want to instantiate standards and just want the dust to settle with concrete examples. Ann’s ‘platform’ and Dale’s transition elements are examples of boundaries we struggle with. Jobst notes that the three interoperability paradigms in messages documents and services, are they all three equal; should there be a preferred paradigm? Are they all equal, or are legacy paradigms preferred or discounted? Is this part of SAEAF or for further down the line, Marc asks? Jobst feels the preferred paradigm approach will reduce costs. If it was only one type of interoperability it might be true, but syntactic vs semantic interoperability differentiate their needs. Ann says we need to make explicit the implemented transactions within their healthcare interoperability community, provide the ability for such community to select a paradigm that meets their needs. What is right for the VA will not work in Wales. End to end preservation of patient information crucial to patient safety.
- Marc notes that we will not solve those things today, still looking for ideas to move forward.
- Charlie McCay notes there is fear uncertainty and doubt for the plans going forward. Some is communication outward to the working group but the other is communication back into the TSC and ArB so we can hear the frustrations early.
- Charles Parisot sees alpha projects all looking forward and would like to see one look at existing artifacts to see how the existing artifacts express the concepts and are not obsolete. Marc notes we must ‘preserve legacy value’; Charlie McCay says he hears the suggestion that we show how our existing value exemplifies SAEAF. Don Jorgenson notes as PASS moves ‘down the stack’ they are looking to pick up existing materials (legacy material) going forward.
- Charlie Mead clarifies that SAEAF is not about services; the directive to create it was explicitly not service focused but to give an entry to services without sacrificing legacy messaging / documents. Some kinds of interoperability don’t require service awareness. If a single event needs more traceability back to the business event you may need the complexity. Service awareness brings rigor to the analysis but does not specify an interoperability paradigm.
- Patrick says they started with existing storyboard. OO harvested past ballot material that got stuck using existing objects to move forward, like Lab.
- Charles also notes on the conformance framework; multiple points of conformance are useful if they are expressed on the business side as well as the technical level.
- Charlie Mead notes the ECCF enables but does not guarantee that conformance statements are made at the different levels. It enables peer-wise conformance assertions about those conformance statements.
- Ann says we have different kinds of standards adopters that have different needs. Vendor wants everyone using same interface, same syntax such as IG profiles. If that succeeds a small national initiative can put out a procurement based on that profile and know confidently they will find it in the market. Have to manage legacy and nudge it towards interoperability with small budget over small rate of change. Need to be able to incrementally implement and not do large scale change. A large managed program can drive large-scale change more easily and all perspectives must be accounted for.
- Charles notes the explanation of the conformance and the means to validate are two different topics. First we must define at a business level what is meant; what you can achieve and conversely what you cannot.
- Mark Shafarman notes the binary statement for conformance, is it English language, true or false, machine language, etc? NCI building an implementable framework. TOGAF is 780 pages. To build out our framework, is it possible to do so on a volunteer basis. Ron Parker notes they’re not implementing the entire framework but a specific implementation of an architecture within the framework. Service contracts go along way towards making the implicit, explicit; and expose the business reason. We need to express how we will use that rigor in evaluation.
- Mark S asks where are the examples we can follow, like from TOGAF, to see from a facilitator sense?
- Marc says we know what the right thing to do is but don’t know how to get there. Once you write it out maybe not elaborated enough to be clear but you’ve got the framework. Now we know we need to build better conformance statements but wow it’s a lot of work. How do we now instantiate that and build their examples. Need to recognize that the problem was always there but the framework pinpoints the parts where we need to elaborate.
- Ron Parker putting together a couple slides on where we are and where we’re going to the next WGM.
- Charlie Mead says the tools NCI is building are independent of the implementation guide; don’t want to build the implementation for HL7 – HL7 must build that.
- Next Steps – 10 minutes
- Flipchart issues:
- RIM alignment/positioning
- ITS: what is a platform? Need to clarify.
- Placement of concepts/artifacts such as RIM& MIF into the framework
- Role of a transforming tool
- Targeting audiences: architects vs standards writers
- Stating a preferred paradigm
- Preserve legacy value & explain/show where it exemplified SAEAF principles
- Conformance: need multiple levels/points, but need to communicate clearly to applicable audiences
- Be aware of stakeholder needs at various levels
|In attendance||Name||Affiliation (if not SAEAF-alpha list members)||E-mail Address||Saeaf-alpha list members|
|x||Constable, Lorraine||Constable Consultingemail@example.com|
|x||Curry, Jane||Tooling WGfirstname.lastname@example.org||x|
|x||Dixon Hughes, Richard||DH4 Pty Ltd/HL7 Internationalemail@example.com|
|Hammond PhD, Edfirstname.lastname@example.org||x|
|x||Hufnagel, Steve||DOD: MHSemail@example.com|
|x||McCay, Charles||TSC Chair, Ramsey Systemsfirstname.lastname@example.org||x|
|x||Mead, Charlie||NCI CBIITemail@example.com||x|
|x||Nelson, Dalefirstname.lastname@example.org, email@example.com||x|
|x||Obayashi, Masaharu||HL7 Japanfirstname.lastname@example.org|
|x||Parker, Ron||Canada Health Infowayemail@example.com||x|
|x||Raup, Gordon||Care firstname.lastname@example.org|
|x||Schwallier, Brian||Vocab WGemail@example.com|
|x||Shafarman, Mark||Shafarman Consultingfirstname.lastname@example.org|
|x||Stechishin, Andy||CANA Sofwareemail@example.com|
|Stevens Love, Helenfirstname.lastname@example.org||x|
|x||Tucker, Mark||Regenstrief Instituteemail@example.com|
|x||Van Dyke, Patfirstname.lastname@example.org||x|
|Van Hentenryck, Karen||karenvan@HL7.org||x|
|x||Vreeman, Daniel||Regenstrief Instituteemail@example.com|
|x||Wrightson, Ann||NHS Walesfirstname.lastname@example.org|