Implementation of a dynamic scheduler for apparel computer integrated manufacturing systems

A major problem faced by the typical apparel manufacturer in day-to-day operations is the need for an effective scheduler to determine actions required because of operator and/or machine non-performance. The dynamic resource allocation system (DRAS) scheduler described in this paper is the natural outgrowth of previously funded research related to the design, development, technology transfer and installation of a CIM system in an apparel plant. The already developed CIM system provides a way for the DRAS scheduler (described in this paper) to change a production process dynamically. The implementation and modeling approach of the scheduler is described. The underlying system model is based on Petri nets and object-oriented databases. Petri nets provide a dynamic model of the CIM system, while object-oriented databases provide necessary information about system components and overall system activity. The data from the CIM system is used by the scheduler in the simulation mode. The daily production plan is used as a guideline to determine how to react to and correct production problems.