Diptendu Sinha

A quantification of the value of flexibility

Abstract Managers of production systems are often faced with the decision of acquiring production capacity to meet forecast demand. Return on investment (ROI) models have traditionally been used, but unfortunate results have occurred in some instances, because the benefits of flexibility were not quantified. This paper concentrates on two aspects of flexibility, the ability to change manufacturing mission and capacity. They are shown to have quantifiable economic value in a particular situation by use of a decision theoretic model. System parameters of interest are investigated to estimate their influence, including uncertainty.

Policy and cost approximations of two-echelon distribution systems with a procurement cost at the higher echelon

This paper investigates a two-echelon (warehouse-retailer) inventory system with stochastic demand and a pull system of inventory allocation. We assume that ordering costs are charged at the warehouse for procuring the item from a supplier. However, the internal costs of ordering the item from the warehouse by the retailers are considered negligible. For this problem, the lower echelon uses a single critical number, an order-up-to-level policy, whereas a (s, S) inventory system is followed at the upper echelon. We develop a cost model for this problem and provide a simple algorithm for estimating the optimal policy. Simulation is used to test the accuracy of the model.

Stochastic analysis of flexible process choices

Abstract In this paper we develop a stochastic linear programming model to evaluate process choice decisions for discrete part manufacturing. We consider both the flexible and dedicated manufacturing processes as well as multiple products at various stages of their life cycles. The model determines the optimal investment schedule in each process alternative that maximizes the expected net present value over different demand scenarios. We also compute the economic value of flexibility and examine the impact of demand variability on net present value.

On asymptotic approximations for (s,S) policies

This paper is concerned with optimal policy approximations based on asymptotic renewal theory, and their accuracy conditions, for a class of periodic-review inventory systems We compare the performances of two such approximations with those of the optimal policies using a wide range of demand distributions and parameter settings Accuracy conditions are derived from new bounds on the optimal policy and from empirical considerations related to the rate of approach of the renewal function to its linear asymptot. Algorithms used in computing asymptotic approximations and optimal policies are also based on new theoretical findings, including a sufficient optimality condition

An inventory decision support system using the object-oriented approach

The paper discusses an object-oriented approach for the construction of an integrated decision support system (DSS). An analysis of computational needs at various levels of an organization defines the data objects of the DSS. The data objects are defined within a hierarchic relationship to support inheritance and aggregation. Using a two-tier approach, we devise high-level abstractions of data and models (system-tier objects) to design the user interface. Next, we consider specific details of implementation-tier objects (as descendants of the system-tier objects) to complete the application. We implement the framework for an inventory management system and provide a simple example to illustrate the use of the DSS.

Expert systems for inventory control management

Abstract An expert system for inventory management is presented in this paper. The focus is on the development of a simple, user friendly tool that can be used by managers to increase the effectiveness of their inventory control systems. The expert system is capable of deriving input parameters by performing necessary analysis on databases, implementing a set of rules for the selection of inventory models, and present the output through tables. plots, or graphs. The scope of this paper is limited to the single item, single location problem.

The design of an expert system for inventory control

Abstract An expert system for inventory management is presented in this paper. The focus is on the development of a simple, user-friendly tool that can be used effectively by managers to increase the cost-effectiveness of their inventory systems. The expert system is capable of analyzing input parameters by performing statistical analyses of data bases, generating plots and graphs, implementing a set of rules for the selection of inventory models, and choosing a solution procedure. The scope of this paper is limited to the single-item, single-location problems.

Knowledge Modularization for Adaptive Decision Modeling

Existing research has shown that theorem proving techniques of predicate calculus can be used as effective tools for query resolution in decision support systems. Implementation of these techniques however require searching large axiom sets leading to dead ends and backtracking. In this paper we use a mathematical programming approach to partition axiom sets into modules for efficient query resolution. The method is based on a two-stage analysis of the domain specific knowledge regarding queries types. The first stage analyzes the queries to determine the information requirement on the axiom set. The second stage determines the optimal modularization of the axiom set using this information. The advantage of our approach is that by selecting the appropriate modules for query resolution only a subset of axioms are considered in the resolution process. This approach works significantly better than the traditional methods when the axiom set can be partitioned into non-overlapping modules based on the commonality of a given query set. In the case of module overlap, the math program assigns the axioms to one of the overlapping modules maximizing overall query resolution efficiency of the system. An adaptive component performs dynamic reorganization of the modules when the query set is time variant. Real time issues of query resolution with modularized axiom set are briefly discussed. INFORMS Journal on Computing , ISSN 1091-9856, was published as ORSA Journal on Computing from 1989 to 1995 under ISSN 0899-1499.

A Structured Expert System for Model Management in Inventory Control

The focus of this paper is on the expert system design for inventory control model management. The model selection procedure uses the knowledge of inventory theory as well as user-specific heuristics and approximation rules. The models available in the model base are represented in a model selection tree using information currently present in the inventory control literature. We explore the use of a structured design methodology through incorporating knowledge modules (i.e., sub-expert system) at the nodes of the model selection tree. The knowledge base is also modularized into selection, application, and solution knowledge bases in addition to a database, each with its unique and independent set of rules and/or storage requirements. This structured design reduces the knowledge representation complexity, facilitates the search process and allows for the use of backtracking procedures to select approximate models for the problem. Implementation of the system using the expert system shell GURU is discussed ...

Optimal policy estimation for continuous review inventory systems

Abstract The paper considers a continuous review ( s , S ) inventory system where the interarrival times of the customers and the quantities demanded by them (demand batch sizes) form two independent and identically distributed sequences. The performance of several easy-to-compute optimal policy approximations based on the asymptotic renewal theory are evaluated. The results are tabulated for a wide range of parameter settings for gamma, log-normal, truncated normal, Weibull and inverse Gaussian batch size distributions. The interarrival time distributions are assumed to be exponential following [1,2]. The exact policy is computed by a search routine using the numerical methods of McConalogue [3,4].