A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Advanced Technologies

Thesis (M. Eng.) - Asian Institute of Technology, 1996

This thesis concerns the development of forecasting, production planning and scheduling models for a pressure container factory in Thailand. The solution procedure for this study is divided into three parts. In the forecasting part, Winter's and Decomposition models are initially used to forecast the seasonal demand. The model with the smaller forecast errors is selected and tested for the remaining of residual patterns. If it can be improved for the accuracy, ARIMA model will be used to resolve and compared with initial forecasting model. The forecasting model which has the smallest forecasting errors is selected. These models are proven as a good forecasting model by statistic testing. The results obtained from the proposed forecasting model are compared with the existing practice of forecasting. In the production planning part, it is a multi-product, multi-period and multi-serial stage with fixed workforce and available overtime problem. This problem is formulated as a linear programming model. The objective is to select production quantities, inventory levels and overtime requirement which minimize the total costs of inventory holding and overtime. In the production scheduling part, four priority rules, namely, run out time (ROT), minimize slack time (MST), critical ratio (CR) and shortest processing time (SPT) are applied to determine sequence of jobs. Then, the sequence is checked for feasibility. If it is not feasible, it will be slightly modified to be a feasible sequence. The priority rule which results in a sequence that satisfy as much customer demand as possible should be selected. Finally, the implementation of the proposed method is made and applied to the production system.