DUE-DATE ASSIGNMENT PROCEDURES WITH DYNAMICALLY UPDATED COEFFICIENTS FOR MULTILEVEL ASSEMBLY JOB SHOPS

This paper presents a study of due date assignment procedures in job s hop environments where multi-level assembly jobs are processed and due dates are internally assigned. Most of the reported studies in the li terature have focused on string type jobs. We propose a dynamic update approach (which makes use of Little's Law) to obtain the coefficients used in the traditional due date assignment procedures of constant al lowance (CON), total work content (TWK) and critical path processing t ime (CPPT). The coefficient assigned to a given job reflects both the state of the shop at the time the job is processed and the characteris tics of the job. The approach also provides the shop management with t he ability to control the average job lateness. In the simulation expe riments conducted in this study, we set the average lateness at zero. The analysis of simulation results shows that the proposed dynamic pro cedures provide overall better shop performance than their static coun terparts, especially for less complex assembly job structures. A proce dure for determining job due dates that extends the critical path conc ept of the CPPT procedure to critical path flow time (CPFT) is also pr oposed. Unlike the others, this procedure does not need the determinat ion of any coefficients. The procedure uses estimates of waiting times at work centers that are determined dynamically based on shop work lo ad information. In this paper, an adaptive adjustment approach is also suggested to bring average lateness for the CPFT procedure to a targe t value. Results of the simulation experiments show that the CPFT comb ined with the adaptive adjustment approach (CPFT-ADJ) provides overall improved performance compared to the dynamic and static versions of t he CON, TWK, and CPPT procedures for less complex job structures. For more complex assembly job structures and string jobs the CPFT-ADJ proc edure results in comparable performance to the dynamic versions of the CON, TWK, and CPPT procedures. The paper also provides an investigati on of the interaction between the the two priority rules: earliest job due date (JDD) and the earliest operation due date (OPNDD) and the fo ur due date procedures: CON, TWK, CPPT, and CPFT-ADJ. In general, for multi-level assembly job structures JDD outperforms OPNDD in terms of average job lead time and tardiness.