Implementation of a Model of Requisition Processing for the Ships Supply Support Study

Abstract

Steps in the Methodology 2-3Step I: Eliminate Cycles from Flow Matrix 2-5Step II: Establish Work Load Factors 2-5Step III: Establish Capacity Factors 2-6Step IV: Implement Multipriority Queueing Model 2-10Step V: Path Analysis 2-10Step VI: Establish Frequency Distribution of Throughput Times Per Path 2-11Step VII: Calculate Distribution of System Throughput Times . . 2-11 Illustration of the Methodology 2-12Step I: Eliminate Cycles 2-12Step II: Establish Work Load Factors 2-13Step HI: Establish Capacities 2-15Step IV: Implement Multipriority Queueing Model 2-16 Steps V and VI: Path Analysis and Path Frequency Distributions. 2-23Step VII: System Throughput Times 2-29 MODELING REQUISITION PROCESSING AT SPCC 3-1 rtStep I: Eliminate Cycles from Flow Matrix 3-3Step V: Path Analysis 3-4Step HE: Establish Capacity Factors 3-7Step II: Establish Work Load Factors 3-9Step IV: Implement Multipriority Queueing Model 3-9Outputs of the Multipriority Queueing Model 3-16Step Generally, the data describing materials handling operations revealed very modest backlogs. Confidence in the throughput times predicted by an extrapolation of materials handling models was enhanced by the fact that the models could be designed to closely replicate the materials handling process. Only that portion of throughput time associated with requisition processing appeared to offer complexities which would not submit to main force analytic techniques. MODELING REQUISITION PROCESSINGStudy requisition processing at inventory control points (ICPs) and stock points revealed long waiting times not apparently related to servicing the requisitions. Certain of these times could be explained by policy and administrative procedures not related to resource levels. However, in the initial modeling efforts waiting times due to backlogs could not be explained by the relatively simple queueing theory models attempted. Several factors could explain the failure to model backlogs:• the queueing models attempted did not take into account the system of priorities under which requisitions were processed,• the resources dedicated to requisition processing are often used in a number of other capacities,• the assumptions about work load arrival rate and service time distributions used by the queueing model might be unrealistic, and• the backlogs might be contrived by supply system personnel for any of a number of reasons ranging from a desire to efficiently use personnel to foster job security. The problem with the status file records as kept was that the prior "history" of a requisition was erased when the file was updated. An investigation by Fleet Material Support Office (FMSO) personnel revealedthat all entries to the status file were retained for 30 days at the inventory control points in a concentrated data format. A very extensive and ambitious programming 1-3 task was undertaken at FMSO to extract these data and generate the outputs enumerated above. A program was successfully completed and a data extraction and data analysis effort was implemented for data describi...