Framework for Prototyping Evolving Manufacturing Control Architectures Z. J. Pasek, S. Marker, B. Tewari Engineering Research Center for Reconfigurable Manufacturing Systems The University of Michigan, Ann Arbor, MI, USA ABSTRACT Control architectures for manufacturing span a broad range of categories from hierarchical at one extreme to purely heterarchical on the other. Such a range covers a broad spectrum of solutions, from quasi-heterarchical structures to self-organizing systems. This paper describes an agent-based framework that provides the tools and components for fast implementation and testing of various control structures in a manufacturing environment. Typical components include, for example, a generic agent for encapsulation of machine-level control, a configurable decision process for the agents, hierarchical or heterarchical control frameworks. Use of these predefined components enables rapid prototyping and instantiation of system-level control. 1. INTRODUCTION To stay competitive, manufacturing companies must develop assets that enable production of goods with top productivity and minimal cost. At the same time these systems must exhibit rapid responsiveness to the frequent and unpredictable market variations. To facilitate these challenges, a new paradigm of Reconfigurable Manufacturing Systems (RMS) was proposed l\l. The RMS approach proposes quick adjustment of production capacity and functionality by enabling rapid structural system changes, both at the hardware and software levels. To achieve the goals of reconfigurable manufacturing, fast and effective integration of software and hardware components has to be facilitated, enabling not only integration at the manufacturing floor level but also its integration with the rest of the enterprise. This paper presents an implementation, development and testing framework that provides the components to rapidly implement any control structure suitable for a manufacturing environment. 2. BACKGROUND Varying market demands, environmental/societal requirements, and continuing progress of technology define a need for manufacturing control systems that are capable of managing changes effectively and efficiently 121. Such control system architectures must be interoperable and capable of handling a wide variety of data models and applications, as well as conforming to ever-increasing demands on performance requirements of process control applications /3, 4/. The hierarchical and heterarchical systems define two ends of the spectrum for control system organization /5, 6/. Hierarchical systems are known to deliver predictable performance, but being based on a set of fixed rules offer little in terms of flexible strategies. The heterarchical systems 161, on the other hand, inspired by biological analogies and market economies, offer improved responsiveness to disturbances, but often at the cost of sub-optimal solutions and lack of predictability. Another class of approaches /7, 8/, holonic control systems, attempts to combine advantages of hierarchy in distributed structure. All of these control architectures are usually designed up front and do not change while in operation. Any modification of the system structure, such as adding/removing a resource (e.g., a machine) or function, usually requires system shutdown. The speed of disruptive external changes, on the other hand, calls for elimination of such events, which eventually leads to 41 Brought to you by | University of Arizona Authenticated Download Date | 6/5/15 7:02 PM