| As product lifecycles become shorter, industry is placing more emphasis on agile manufacturing systems---systems with unparalleled flexibility. Recent years have shown a shift in paradigms from exploiting economies of scale to economies of scope. The ability to quickly adapt to changing market forces, new product designs, and changing technologies is now key to an organization's survival.;For years now, industry has lead academia in the industrial automation field. Heuristic processes for designing conventional hard automation have been created through trial and error. A formal design methodology for creating such systems is still lacking even today. The problem is exacerbated when these ad hoc methods are applied to agile systems which are vastly more complex. As a result, agile systems, as well as all but the simplest conventional automation systems are prone to run over-budget, produce both over and under designed components, are difficult to debug, provide for little component reuse, and are extremely difficult to maintain and upgrade. A new design process is needed.;This thesis introduces a user-orientated UML-based design methodology for integrating large-scale agile automation control systems projects. The methodology is composed of a UML-based notation (with new extensions) and a design process. A component-based reference architecture truly adapted to the real needs of the industrial environment and suited for different cell configurations is developed. New metrics will be provided that allow the user to analyze the overall flexibility, modularity, and complexity of both the early and final design.;Two applications where this design methodology was implemented will be examined in detail. In the second case, it will also be shown to be applicable to conventionally controlled manufacturing systems as well. With the proposed approach, many of the problems typically seen during agile control system development were significantly reduced. |
