Discrete Event Simulation of Biological Control Processes and their Application to Autonomous Decision-Making in Manufacturing Systems

Lead Research Organisation: De Montfort University
Department Name: Engineering


The proposed project will determine the feasibility of developing a new approach to operations design, planning and control which is based on the recognition that both design and planning activities must be an integral part of the physical control activities. Such methods already partly exist, e.g. Kanbans provide physical controls but with limited planning capabilities and within a limited range of manufacturing system types. However, such integral design, planning and control environments exist within organic cells in that a wide variety of cell products are able to be manufactured, assembled and distributed using purely physical control devices, i.e. with no requirements for manual control intervention or computer based design and planning tools such as assembly line balancing, work-to-schedules and resource requirements plans. The feasibility of autonomous design and planning of 'manufacturing operations' becoming an integrated part of the autonomous control processes taking place within the cell is, therefore, worth exploring because of their potential in significantly reducing the level of manual intervention currently required. Within the intended research, discrete event simulation (DES), extensively used in the planning and design of large and complex manufacturing operations, will be essential in understanding the roles the novel biological control principles can play within manufacturing systems.DES is widely used and sufficiently flexible to provide accurate and detailed models of a wide range of industrial and service operational systems. It enables the significance of individual 'low-level' control actions on the overall system 'functional' performance to be observed and quantitatively measured. Such is the capability of DES that its use may help to resolve a major gap in the functionality of existing systems biology modelling and simulation tools, ie existing tools are unable to show the 'control' significance of biochemical reaction networks which are generally described as chemical reactions. In this low-level, non-hierarchical, notation, symbol nodes represent potential states of molecules or molecular assemblies, and arrows represent state-transitions. State transitions are events, and are assumed to occur spontaneously (i.e. without being triggered), with a probability that is characterized by the average number of transitions per unit of time that would be observed under static conditions. Each node has a population of molecules, and each member of the population can make the state transition independently, forming, at large populations, a massively parallel system. The chemical reaction modelling framework yields a description in strictly thermodynamic terms, in which functionality and hierarchy are only perceived through human eyes. However, without an interpretation of the system in terms of function and hierarchy, it is just as difficult to grasp the 'control' significance of a simulation result as it is to assign meaning to an un-annotated DNA sequence. Additional modelling functionality, for inclusion within the DES model of the bacterial flagellar assembly, will be developed that demonstrates the feasibility of integrating biological processes at the molecular mechanistic level with those at the (apparent) control level. During the research project close collaboration between manufacturing and system biology researchers will exist with exchange of knowledge and expertise between the two research disciplines. These conditions will provide an ideal environment for examining new perspectives / that of manufacturing systems / to systems biology that could aid the generation of hypotheses on the control of intracellular 'building projects'.


10 25 50