On Modeling and Implementing Shopfloor Control Systems

Doctoral thesis, 1997

One of the key factors that hold shopfloor control (SFC) software evolution back is the exorbitant cost, time, complexity, and inflexibility of developing customized SFC software solutions. The development and maintenance life-cycle of SFC software is characterized by lock-in problems. It is a fact that shopfloor environments constantly are changing, while the technology surrounding and supporting them often is changing faster. To handle these changes, the necessary software evolution has become the bottleneck. In this thesis, a literature survey shows that there exists a uniform view of general requirements on generic structures for shopfloor controllers. This work has adopted these general requirements in the design for a shopfloor control concept named Production Activity Control (PAC++). To gain experience and knowledge about modeling requirements, three pilot cases were defined. The objective in each pilot case is to gain an understanding of the development and implementation phases of shopfloor control models. The cases were placed on the work station-, cell-, and shop level in the National Institute of Standards and Technology (NIST) hierarchical architecture. The PAC++ concept shows that by using a consistent modeling and implementation formalism, software modules become distributed and scalable. This gives the advantage to start with small and well-defined projects, while retaining the possibility to use an incrementally growing approach to a full-scale Shopfloor Control System, SFCS. Finally, the PAC++ has the built-in flexibility to improve and enhance its functionality during the whole life-cycle of a shopfloor control system, justifying development cost and time.