Manufacturing control systems: An approach to reducing software costs

Abstract

Real-time control systems are expensive to develop and few of them operate properly when first delivered. Much of the development cost can be attributed to errors which are made in the early phases of the software development. These are often the hardest to detect and the most expensive to correct. The use of explicit models to test the consequences of design alternatives holds promise for reducing this problem. Simulation of both the physical and control systems provides the explicit models necessary not only to detect logical errors but also to examine the complex, dynamic relationships of control design decisions. Simulation with the “Activity Cycle” approach uses directed graphs to depict system logic for ease of comprehension and provides a one-to-one correspondence between the simulation and the desired manufacturing control system software. The use of Activity Cycle based design and simulation methods has the potential to reduce substantially the risks and costs involved in the construction of manufacturing control system software.