DEVELOPMENT OF AN EMPIRICAL MODEL FOR JOBSHOP FLOW TIME AND DUE-DATE PREDICTION

Abstract

A computer software was developed and tested for validation for jobshop flowtime and due date prediction. The software utilises data from shop floor which are analysed to predict expected waiting time on each individual machine on a periodic bases. Projection of these waiting times are made for future flow time prediction by using moving averages. The methodology employed implicitly accommodates job characteristics, shop status, system characteristic as well as delaying factors external to the shop. This software is capable of prescribing a range of due-dates to reflect the best, most likely and worst case possible scenarios with their times. When the total theoretical time of a job was 570 minutes, with the arrival date and time 24/08/98 and 8:00 hours respectively the software was able to predict the due-dates for best case scenario, most likely scenario and worst case scenario thus: Best case Scenario due-date: 24/08/98 Time: 12:08 hours Most likely Scenario due-date: 25/08/98 Time: 12:08 hours Worst case Scenario due-date: 26/08/98 Time: 13:08 hours This range will greatly assist management in negotiations with prospective customers. Improving performance of job shops and their ability to execute jobs within due-dates.