The objective of model-driven development is to provide a common point of reference to the developer, the user, and the development tools.  The model becomes the application.  Tools transform the application from a specification form into an executable form for a particular deployment environment.  By being model-driven, specifications are visible and shared, reducing the overhead of coordination. 

In model-driven development, it is important to ensure that the systems being designed are well grounded in the business and that business operations take best advantage of the technology to improve competitiveness.  UML Models and business process models are used to reach both goals.  They are used initially to identify processes that transform business events into data.  And, if desired, this same process model can be used to reengineer these business processes, resulting in a new business process model. 

UML Class Diagrams and Entity-relationship data models have been used extensively to represent the business requirements of organizations.  The goal of model-driven development is to quickly produce a quality data model that accurately reflects the business.  In order to meet this goal, the concept of primary data is used.  Primary data represents the events and activities of the organization, which must be accurately represented if the enterprise is to continue to operate. 

Primary data is available in all organizations within business forms, computer panels, manual or computer file layouts, and other sources.  The data specifications in these sources can be recovered and used to automatically produce a first cut Conceptual Data Model using the appropriate reverse engineering and expert-assisted data modeling tools. 

The data model produced from the primary data is enriched to become an object model or Application Data Model.  One of the important characteristics of an object model is the close integration of the data and the rules governing the behavior of the data.  Older methodologies required a dual modeling of data and of the processes operating on the data.  In the object model it is possible to encapsulate the data with the operations that act upon it and with the business rules governing those operations. 

To ensure success, the approach used by Astron Technologies Inc. relies on the ongoing participation of all involved parties in the design and design validation process.  The use of a cross-functional project team as the decision making body during each phase ensures understanding, cooperation, and ownership throughout the development cycle.

[Up] [Principles] [Objectives] [Disciplines] [Overview]

Copyright © 1999 - 2004 Astron Technologies Inc.