Discrete Event Simulation of Biological Control Processes and its Application to Autonomous Decision-Making in Manufacturing Systems |
Scientific & Technical Relevance
To remain competitive there is an increasing move in industry towards mass customisation [1] that is leading to customer demands for greatly increased levels of product and process variety, reduced demand for individual products, shorter product life cycles and significant decreases in supply lead times. All such changes are resulting in increasing amounts of data being required to design, plan and control industrial operations and, despite the increasing use of computer technology, higher levels of manual intervention in design, planning and control decisions. However, improving productivity levels and production responsiveness to such demand changes requires the removal of lead time delays and errors caused by human intervention such that decision-making becomes autonomous in response to changes in competitive market environments.
Such levels of autonomy are increasingly being required since the efficiency of existing techniques used to design, plan, schedule and control manufacturing has always been constrained by their lack of effectiveness in complex environments such as those that exist in high product variety/low demand volume (HPV/LDV) and extended supply chains [2]. Methods that optimise do so only where manufacturing system size is small, and product and demand variety is low. Increases in the complexity of the environment cause these methods to provide efficient solutions to planning problems and rapid response to changes is possible. Non-optimising heuristic and priority based methods are, hence, normally employed because of their speed of response.
Little has changed in this respect for several decades despite the addition of advanced modelling tools such as genetic algorithms. Even the effectiveness of these methods quickly falls when levels of complexity and variability increase. Until recently the response of manufacturing has been to seek to reduce levels of variability within the manufacturing environment as witnessed by the development of cellular manufacturing techniques and more recently lean practices. However, in recent years the growing move towards mass customization is quickly offsetting the benefits once derived from use of these ideas. Because of these increases in complexity and speed of response, therefore, a significant gap is appearing between the design, planning and scheduling requirements of modern manufacturing systems and the ability of existing operations design and planning techniques to meet these requirements.