Automated Verification and Generation of Flexible Automation Control

Doctoral thesis, 2007

Consumer product life-cycles are constantly shortening; the automotive industry is an illustrative example. As a consequence, the introduction of new products into the manufacturing system necessarily becomes more frequent. Inherently, this brings a performance reduction for the manufacturing system. The reduced performance is caused by a down-time and a ramp-up-time. During the down-time the mechanical equipment is rebuilt and the new control programs are debugged. During ramp-up there are a large number of errors mainly caused by mechanical devices not being properly adjusted, bugs in the control programs and operators not used to new procedures. Thus, in order to maintain the productivity level and to achieve full cost-efficiency both the down-time and the ramp-up time must be reduced. One way to reduce these lead times is to verify the control programs in offline mode. However, efficient and reliable offline verification requires some major improvements of the current development process of manufacturing systems. Information handling and development of control programs based on information reuse are the two most important improvement areas. The work presented here addresses four industrial problems related to this, lack of tools for offline verification of control programs, lack of information reuse in the development process of a manufacturing system, lack of operator support in error situations, and lack of tools for analyzing the control of complex manufacturing cells. We propose a development method where information from different tools in the development process of a manufacturing system is reused and processed by tools for verification and optimization. Then the control programs are generated by combining the processed information with a library of standardized software components. The proposed method solves the above-mentioned industrial problems without adding work to the development process. On the contrary, the amount of work will be reduced since the control program development will be automated and the time for debugging the control programs on the shop floor will be drastically reduced, due to the new mathematically based verification process.