This paper describes the use of and future development strategy for a simulation-based Critical Resource Scheduler for batch manufacture. The need for continual research in Production Planning and Control in batch manufacturing is created by the contradictory conditions of a multi-goal, complex, and dynamic problem domain for which simple and understandable solutions are being sought. Operations Research-based optimisation approaches using linear and dynamic programming have been formulated and refined but have been unable to conduct timely searches of the vast solution domain. The constraints on use are problem size capability, the dynamic nature of the manufacturing problem, and an inability to handle the multi-goal criteria of interest. Simulation has proved to be a feasible method for evaluating possible scheduling solutions.
Commercial computer packages have been designed for rapid modelling of manufacturing scenarios. Use of this type of proprietary simulation package is predominantly to support the justification of new facility investment and is not generally used for on-going manufacturing control. Purpose-coded models have greater potential for on-going use in production planning and control.
The aim of the research described in this paper is the design and implementation of a modular simulation package for applied industrial control. The key features are the intended user base (small- to medium-sized companies), methods to identify the critical resource, and methods to use the critical resource as the focus for the sequencing of parts.
Designated MSA (Manufacturing Simulation and Analysis), the simulator suite of programs uses one modelling entity, the "Universal Transfer", to represent all manufacturing resources. This enables flexibility in the modelling approach and more user-friendly software.
This paper begins with a brief review of the important aspects of Version 1.0 of the MSA approach: objectives, methodology, implementation, some results, and limitations.
The conclusion of the paper details Version 2.0 of the MSA methodology, focusing on the planned object-oriented implementation of the simulator. A paper by Driscoll and Hurley  contains a more in depth description of the results than is presented in this paper.