MODELLING AND SIMULATION OF INVENTORY SYSTEM IN ARMY WORKSHOPS

Abstract

The diversity of terrain and usage pattern depending upon the employment and deployment of. Army equipments make it extremely challenging to provide them an effective repair and maintenance cover. Availability of adequate spares is the primary requirement for carrying out any repair work. The random nature of equipment failure and unusual deterioration in operational environments are few factors causing accelerated wear and tear. Planning and maintaining an appropriate inventory level of spares, therefore, is something that always remains the matter of great concern for.the OsC field / station workshops. The work, as applicable to Army Workshops, has been completed in following steps: - (a) Critical analysis of existing inventory policy. This highlights the inadequacy of procurement policies based on MMF(Monthly Maintenance Figure), PAF(provisioning action figure) and a global lead time value irrespective of operational commitment and the place of deployment. (b) Analysis of demand and lead-time patterns for selecting suitable distributions for demand and lead-time simulation. (c) Development of a Dynamic (r, Q) model with stochastic Demand Pattern and Lead Time. The model takes into account the penalty cost arising due to stockout and excess of payment while resorting to local purchase. (d) Derivation of expected number of stockouts for different demand distributions such as: - (i) Normal Distribution (ii) Exponential Distribution (iii) Poisson Distribution (iv) Weibul Distribution (v) Gamma Distribution The derivations have been carried out for the following three options: - (i) Fixed Service Levels (ii) Known ROP (iii) Fixed Maximum spending on stockout (a function of the extent of local purchase) (e) Derivation of a mathematical relationship, which gives annual stockout cost and limits for local purchase. (f) Mathematical derivation of fixed service level for an Army Workshop. (g) Analysis of economic implications of the policy and computation of expected savings. The developed model has been simulated for the lead-time and wastage pattern using a PC. The simulation software, developed on Microsoft Visual Basic platform with Excel spread sheets as database, uses these patterns to optimize the order quantity and the reorder point and compares the proposed model with the existing one in terms of incremental costs. To represent the demand during lead-time, two types of probability distributions, Normal (if the monthly demand exceeds 20) or Poisson (if the monthly demand does not exceed 20) - as specified by EME Training Advisory Committee in the terms of reference - have been considered as important aspects of inventory simulation. The option of performing chi-square testing has also been incorporated for suitable curve fitting of the monthly demand pattern. This option also takes into account Exponential demand distribution in addition to Normal and Poisson. To verify the impact on the overall inventory cost, the model has been applied on a large population of class A