Operation Specification for Sequence Planning and Automation Design

Sammanfattning: When designing an automation system, an important challenge is to specify the different tasks, or operations, of the system. This has become a problematic bottle-neck in industry, since the rising needs on manufacturing flexibility increases the complexity of the entire manufacturing system, as well as its control function. This thesis studies how to specify the manufacturing operations when developing an automation system, especially how to plan the operations and the relations among them, called sequence planning, and how to implement them in a programmable logic controller (PLC). To manage the complete automation development process, a formal language called sequences of operations, SOP, is introduced. This language, together with a set of methods, are used to represent, identify and visualize complex operation relations, so that the behavior of the system can be understood. By the use of different views or perspectives, the sequence of operations related to e.g the part flow, robot operations or operator tasks can be visualized. These different perspectives help the user to easier understand the relation between the control of a cell and the mechanical design, as well as the product design and the total system behavior. The proposed language is based on a formal model, thus formal verification, optimization and synthesis are available. This is necessary since it is hard to find errors in complex system specifications. When implementing the control of manufacturing operations, much of the code is related to supporting concerns like alarm handling, logging, user interface, etc. These parts tend to be hard to fully reuse and implement, since they are tangled with other functionalities, and are located at many places in the program. An aspect-oriented programming (AOP) approach is proposed, to reuse and implement these concerns. AOP is an emerging programming methodology in computer science, which helps the programmer by separating crosscutting and tangled concerns in a program. The approach is focused on how to adapt the AOP methodology for PLC program development, and how it can benefit the control design process.