Monday, October 19, 2009

Integration Points

An integration point between systems is the point where value is moved from one system to another. Understanding and creating the integrations where the value successfully transfers between the systems is often quite difficult. Successful value transfer even between simple systems, such as two humans trying to do knowledge transfer, can be quite complicated.

When analyzing systems of systems it is useful and necessary to collect information about these transfer points. I find the most interesting and useful piece of information about integration points is the class (type) of the integration point. I like to classify the integration points as Human, Custom, Shared, Service, Linked, and Modeled.

A human integration point is one where the value is transferred by humans operating within human systems.

A custom integration point is one where the value is transferred by creating a customized transfer process.

A shared integration point is one where the value is placed into a central repository.

A service integration point is one where value is transferred by a service that is delivered by a tool system.

A linked integration point is one where dynamic, extensible data is delivered by a tool system.

A modeled integration point is one where value is delivered by collaborative models.

It is interesting to look at the characteristics of each of these types of integration. The following table takes a brief look at some of these characteristics. My apologies that this table is difficult to read. You can find a more legible copy on our web site www.hewettresearch.com. Since it isn't as well known, I'll write more about modeled integration next time.

4 comments:

  1. The table is helpful to me as I am example driven. I'd like more information about some of the row headings, such as what you mean by "Entity Identification" and "Agents"
    It'd also be great to hear more about "extraordinary model/EX"
    I couldn't find this table on your website, BTW.
    Looking forward to your next entry...

    ReplyDelete
  2. Interesting abstraction of system communication and intelligence. Software systems gathering knowledge and publishing that in a format that is consumable by other systems, subscribing to services and functionality. In human terms, broadcasting my interest in a topic that someone in the global network knows something about. The obvious, concrete interaction is so inline with SOA; however, this concept reaches much farther beyond publish and subscribe. The integration of knowledge involves AI and learning in order for systems to improve. Wonderful post. Thank you.

    ReplyDelete