If you want to build a "service oriented" enterprise-management architecture, I encourage you consider the following steps. More details on GEM and EMA are posted at http://www.one-world-is.org/rer/owis/publications.htm. The methodology page I refer to is specifically at http://sites.google.com/a/one-world-is.org/gem-ema/Home/gem-approach. I've also posted my collection of several hundred annotated PowerPoint slides that I've put together over a few decades that illustrate the above methodology and steps. They are listed as the 2003-2008 updates, at http://www.one-world-is.com/rer/owis/gem-core/GEM_Core-Collection.htm. If you feel that this information is useful to you, please consider making a financial contribution to my efforts here. A. For SOA, consider that all the "services" of a SOA effort are "information products and perhaps packaged processes" as "outputs" from some activity within a procedure. These information products/packages have metadata/rules, that contain data, that has a specific agreed-meaning to both the producer/sender of the information product/package and the user/receiver. You need the semantics of the information product/package's purpose before you can formalize the data and metadata. Section 2 of the GEM Methodology covers this. B. But before you can get to the semantics, you have to know "who" the sender and receiver are that have to agree, perhaps as persons, but definitely as roles, held by a positions, within offices (staff, program, or project office) or teams, within organizations, at specific physical (geospatial, postal), virtual (IT network, telephone network, perhaps radio or satellite network) locations. Section 3 of the GEM Methodology covers this. C. If you need a repository to store this rich collection of multidimensional information from this enterprise modeling effort, I recommend an extended metadata repository tool that enables you to build this enterprise model all the way out to a full ontology, and thus knowledge-base of the organization, and then to a full axiology, providing a semantically-enriched value-chain, or economy, model. The "technology standards" for the capabilities you'll need to build/buy in order to do the above modeling and to actually implement these information products/packages as a software system, residing on a computing platform, operating over a compute network, is covered in Section 1 of the methodology. You might find the order of the EMA methodology Sections to be inverted, but most organizations I know of have started with Section 1 first, with their inventory of existing IT. You're starting with modeling the domains, which has nothing to do with IT, but everything to do with communication and cooperation across: 1-locations, 2- organizations, 3-organization units, 4-functions, and 5-processes, using the 6-resources of "information products/packages" as your 7-required-capability's process input, control output, and mechanism, and you'd design these specific resources using the Section 2 guidance cited above. If you read the order of these seven "classes" or types of enteprise subjects, you'll see they provide an orderly, effective, and efficient structuring of your enterprise (an enterprise model, architecture, and ontology) to support top-down BPMN. Then you'll need to explore and decide on the business process management (workflow) and data-sharing/ messaging capability (manual or automated) you'll need to develop, test, implement, and deploy it for operation. Roy Roebuck +1 (703) 598-2351 |
|
|