Multi-echelon Inventory System Research Articles

Optimization of multi-echelon spare parts inventory systems using multi-agent deep reinforcement learning

Multi-echelon inventory systems are commonly used in practice to satisfy widely distributed random demands of spare parts in an efficient and cost-effective manner. Optimization of a multi-echelon inventory system is a decision-making problem under uncertainties. Classic inventory policies (e.g. (s, S) and (R, Q)) that do not consider the inventory positions of other warehouses become suboptimal due to interrelationships among different warehouses caused by transshipment. The Markov decision process (MDP) is an effective tool for inventory optimization, which does not require a predetermined parameterized policy structure. Unfortunately, both the state and action spaces of MDP suffer from the curse of dimensionality when the number of warehouses increases. This paper optimizes the inventory of a large-scale multi-echelon inventory system using a new multi-agent deep reinforcement learning (MADRL) algorithm named EM-VDTD3 that is developed by introducing value decomposition and experience buffer modification into the twin delayed deep deterministic policy gradient (TD3) algorithm. Each agent in EM-VDTD3 manages a subsystem in the multi-echelon inventory system. Because different agents share the same network parameters, networks are customized to process subsystems with different parameters. Domain knowledge of inventory control is embedded in the learning process of EM-VDTD3 by adding expert experiences to the experience buffer. An efficient approximate method is developed to identify a teacher policy that generates expert experiences. Numerical studies about a spare part inventory system in the wind energy industry show that the proposed EM-VDTD3 outperforms benchmark methods.

Reliability-Driven Multiechelon Inventory Optimization With Applications to Service Spare Parts for Wind Turbines

There are multiple warehouses in a multiechelon inventory system, and the size of the state space increases exponentially with the number of warehouses. Therefore, the curse of dimensionality becomes unavoidable when performing steady-state analysis. Most existing studies calculate the inventory cost or supply chain reliability based on specific assumptions. For example, it often assumes that the lead time is either zero or an integral multiple of the review period, and that each warehouse adopts a base-stock policy. This article considers a more practical and prevalent situation where the lead time is less than a review period, and a more general (s, S) strategy is adopted. The curse of dimensionality during steady-state analysis is alleviated by decomposing transition probabilities. Then, the cost and supply chain reliability are derived from steady-state distributions. Finally, a case study involving spare part inventory of wind turbines is considered. Nondominated inventory strategies are obtained using the particle swarm optimization method to strike a balance between costs for the wind turbine manufacturer and wind farm owners.

Simulation-Based Approach for Multi-Echelon Inventory System Selection: Case of Distribution Systems

Due to the current complexity of the supply chain, multi-echelon inventory management has become challenging while also being an interesting field of research as it allows efficient control of supply chain interdependencies. It became clear to many researchers that analytical models are no longer effective for addressing the multi-echelon inventory management problem. Simulation can be used to assess and quantify the impact of each inventory strategy on a supply chain performance. Our paper aims to provide a simulation-based approach to guide decision makers select and validate a multi-echelon distribution inventory system. The proposed approach is composed of four major steps that involve characterization of the current supply chain, conceptual modeling of the multi-echelon inventory system alternatives, and finally, simulation modeling using appropriate simulation software to compare and test different options. The approach was also tested and validated through an application to the case of the Moroccan pharmaceutical products supply chain in the public sector. The results of the simulation demonstrated that adopting an installation stock policy at all levels of the supply chain with an allocation of safety stocks in the most downstream stages is the best and most appropriate alternative for the pharmaceutical supply chain under study.

Multi-dimensional classification of wind turbine spare parts in a multi-echelon inventory system

Multi-dimensional classification of wind turbine spare parts in a multi-echelon inventory system

Multi-Dimensional Classification of Wind Turbine Spare Parts in a Multi-Echelon Inventory System

Multi-Dimensional Classification of Wind Turbine Spare Parts in a Multi-Echelon Inventory System

An order-splitting model for supplier selection and order allocation in a multi-echelon supply chain

This paper considers an integrated supplier selection and inventory control problem for a multi-echelon inventory system with an order-splitting policy. A buyer firm consisting of one warehouse and N identical retailers procures a type of product from a group of potential suppliers; the acquisition of the warehouse takes place when the inventory level depletes to a reorder point R, and the order Q is simultaneously split among m selected suppliers. We develop an exact analytical model for the order-splitting problem in a multi-echelon system, and formulate the supplier selection problem in a Mixed Integer Nonlinear Programming (MINLP) model. This model determines the optimal inventory policy that coordinates stock levels between each echelon of the systems while properly allocating orders among selected suppliers to maximize the expected profit. For verification and validation of the proposed mathematical model, we conduct several numerical analyses and implement simulation models which helps us demonstrate the model’s solvability and effectiveness.

Performance-based inventory classification methods for large-Scale multi-echelon replenishment systems

Inventory classification is a managerial method utilized to group items that share predetermined characteristics, with the intent of assigning group-specific controls and monitoring mechanisms to each established item group. In this paper, we develop a performance-based inventory classification (PBIC) method that finds a grouping solution for a multi-item, multi-echelon inventory system controlled by continuous review. We argue that instead of grouping items based on similarities in unit cost, demand rate, or leadtime, the most effective strategy is to group items based on the information contained in their control policy values and their performance-related parameter values. We introduce a new policy-driven approach for establishing the classification criteria used to group items. We also adopt a ranking method to control the multi-dimensionality of multi-echelon systems in order to determine a one-dimension score. To group items, we improve the Pareto-based (ABC) solution by developing a search-based partitioning solution, utilizing a novel aggregation process. Our findings indicate that the PBIC method significantly outperforms alternative classification methods. Also, the empirical results show that there is a negligible gap between the performance of the PBIC and the optimal (complete enumeration) grouping solution. Finally, we discuss our work in the context of managerial implications highlighting the use of classification for problem aggregation and size reduction, when managers need to perform efficient, yet extensive, and dependable what-if analyses related to inventory management.

An agent-based approach for resources’ joint planning in a multi-echelon inventory system considering lateral transshipment

A joint planning of spare parts and maintenance workers in cross-echelon and lateral transshipment can reduce total system cost; however, only a limited part of the literature considers this combination. In this paper, we propose an agent-based generic evaluation model for the joint planning of resources by examining the multi-echelon inventory system with the operation management of multi-component equipment. First, a dual-clock failure modelling mechanism is provided to accurately generate the maintenance resource requirements. Afterwards, a resources scheduling process based on the contract net protocol is described. In addition, the dynamic change in the spare parts inventory is also analysed by considering whether the replaced failure components can be repaired or not. Moreover, a system cost assessment method is developed to compare different scheduling strategies and to select the one that can minimize the expected total cost. Finally, we examine an example of a two-echelon inventory system to verify the effectiveness of the proposed approach by using four representative scheduling strategies: fixed-pair, nearest distance, maximum inventory and minimum ratio. The results indicate that the approach can well support the joint planning of maintenance resources.

An Optimization Model for Multi-Echelon Multi-Commodity Supply Chain Design with Side Constraints

A proper strategies in designing and optimizing supply chains can help enterprises to stay competitive in this volatile economy. Identifying various type of products that run through the supply chain and aligning several parts of the supply chain to these product types are considered to be effective strategies adapted by successful companies. In this paper we address an optimal design of a multi-echelon supply chain problem with various commodity that is of significance to the business enterprise, along with the associated inventory systems. By using the guaranteed service approach to model the multi-echelon inventory system, we develop an MINLP model to simultaneously optimize the transportation, inventory and network structure of a multi-echelon supply chain which involve multi commodity. The optimization model is then solved using a strategy of releasing nonbasic variables from their bounds, combined with the “active constraint” method.

Simulation-based optimization of control policy on multi-echelon inventory system for fresh agricultural products

As fresh agricultural products are perishable and vulnerable, reducing inventory cost is a strategic target for supply chain enterprises. How to design a reliable multi-echelon inventory control policy is still a great challenge. Therefore, the inventory cost of a three-level fresh agricultural products inventory system was firstly mathematically analyzed. Then, the simulation-based optimization model of the multi-echelon inventory system for fresh agricultural products was proposed by using the Flexsim simulation software and the improved particle swarm optimization algorithm. Finally, the multi-echelon inventory system is simulated based on a large number of survey data. Simulation results demonstrate that the proposed simulation-based optimization model of multi-echelon inventory system for fresh agricultural products can provide decision-making and technical support for the formulation of inventory control policy, and also it shows that the modeling of system simulation is an effective method to solve the problem of complex system. Keywords: multi-echelon inventory system, simulation-based optimization, fresh agricultural products, control policy, Flexsim simulation software DOI: 10.25165/j.ijabe.20191202.2834 Citation: Xu G Y, Feng J H, Chen F L, Wang H, Wang Z F. Simulation-based optimization of control policy on multi-echelon inventory system for fresh agricultural products. Int J Agric & Biol Eng, 2019; 12(2): 184–194.

Simulation-based optimization of control policy on multi-echelon inventory system for fresh agricultural products

Simulation-based optimization of control policy on multi-echelon inventory system for fresh agricultural products

A literature review on multi-echelon inventory management: the case of pharmaceutical supply chain

Inventory management remains a key challenge in supply chain management. Many companies recognize the benefits of a good inventory management system. An effective inventory management helps reaching a high customer service level while dealing with demand variability. In a complex supply chain network where inventories are found across the entire system as raw materials or finished products, the need for an integrated approach for managing inventory had become crucial. Modelling the system as a multi-echelon inventory system allows to consider all the factors related to inventory optimization. On the other hand, the high criticality of the pharmaceutical products makes the need for a sophisticated supply chain inventory management essential. The implementation of the multi-echelon inventory management in such supply chains helps keeping the stock of pharmaceutical products available at the different installations. This paper provides an insight into the multi-echelon inventory management problem, especially in the pharmaceutical supply chain. A classification of several multi-echelon inventory systems according to a set of criteria is provided. A synthesis of multiple multi-echelon pharmaceutical supply chain problems is elaborated.

A Comparison Study of Stochastic- and Guaranteed- Service Approaches on Safety Stock Optimization for Multi Serial Systems

Two competing approaches have been developed over the years for multi-echelon inventory system optimization, stochastic-service approach (SSA) and guaranteed-service approach (GSA). Although they solve the same inventory policy optimization problem in their core, they make different assumptions with regard to the role of safety stock. This paper provides a detailed comparison of the two approaches by considering operating flexibility costs in the optimization of (R, Q) policies for a continuous review serial inventory system. The results indicate the GSA model is more efficiency in solving the complicated inventory problem in terms of the computation time, and the cost difference of the two approaches is quite small.

An Integer Programming Model for Multi-Echelon Supply Chain Decision Problem Considering Inventories

In this paper we address a problem that is of significance to the industry, namely the optimal decision of a multi-echelon supply chain and the associated inventory systems. By using the guaranteed service approach to model the multi-echelon inventory system, we develop a mixed integer; programming model to simultaneously optimize the transportation, inventory and network structure of a multi-echelon supply chain. To solve the model we develop a direct search approach using a strategy of releasing nonbasic variables from their bounds, combined with the “active constraint” method. This strategy is used to force the appropriate non-integer basic variables to move to their neighbourhood integer points.

Modelling and optimisation of multiproduct, multi-echelon inventory problem

The purpose of this paper is to determine the optimum lot-sizes and re-order intervals of a 3-echelon supply chain system. The mathematical model is built based on Roundy's PO2 policy and integer policy of ordering as a constrained non-linear programming problem. For illustration purpose, we took two problems of single and fifty products distribution systems under deterministic condition. Problems are solved with exhaustive search method on spreadsheet and through Matlab programming. Though PO2 policy is very simple and is able to provide a few solutions, and faster, many times it fails to find an optimal solution and sometimes, any feasible solution at all. On the contrary, integer policy gives many including optimum and all PO2 solutions. Result shows that our proposed model and the simple algorithm applied for the solution have superiority and is effective on reducing the total cost of the multi-product, multi-echelon inventory system. Further, the products are grouped based on reorder interval using joint replenishment strategy.

Economic Design of Multi-echelon Inventory System with Variable Lead Time and Service Level Constraint in a Fuzzy Cost Environment

Normally, human based industrial decision making involves both quantitative and qualitative input factors such as ordering cost, setup cost, etc. Thus the practitioners should be careful in accounting flexibility in the input factors. This paper optimizes an inventory system by employing mathematical model with variable lead time and service level constraint (SLC) in a fuzzy cost environment. Fuzziness is established by allowing the cost components imprecise and vague to certain extent. Trapezoidal and triangular fuzzy numbers are used to represent these characteristics. We also use signed distance method to defuzzify the fuzzy joint total expected cost and differential calculus optimization technique is adopted to find all optimal solutions to the model. Numerical results highlighting the sensitivity in the decision variables are also described.

A simulation- and genetic algorithm-based optimisation of closed-loop multi-echelon inventory system

The objective of this research is to develop a genetic algorithm (GA)-based optimisation approach for a multi-echelon closed-loop inventory system of items, which are repairable in nature. In the context of the passenger transportation industry, engineering aggregates like engines, alternators, axles and tyres are representative examples of such systems. The present work is motivated by a real-life example of state-owned transport corporation having more than 9,000 buses. Operationally, the corporation is divided across several depots (the lower most echelon) and divisions (the next higher echelon). The contribution from this research is manifold. In terms of specific insights, it is established that keeping a higher base stock of spare tyres at the divisions than at the depots is operationally better. This is a consequence of the risk pooling effect. The present research enables us to understand the optimal policy parameters of a complex multi-echelon inventory system. The whole optimisation approach and the simulation model can be generalised and can fit well in several other related problems and their contexts. Typically, the approach is applicable to the problems of repairable-parts inventory related to industries with heavy utilisation of equipments like the chemical and the petrochemical industries.

Repairable spare parts flow in a multi-echelon inventory system

This paper considers the problem of changing the value of selected characteristics for the operation of a two-level inventory system of maintainable parts depending on the flow of regenerated components in the inventory system. We analyzed the influence of the flow of regenerated components on the value of the coefficient of peripheral storage fill-rate, total costs and the average waiting time in the queue in peripheral warehouses, at different nominal values of the stocks in central and peripheral warehouses. The demand for spare parts in the peripheral warehouses is dependent on seasonal changes, modeled with an appropriate probability distribution.

A continuous-time multi-echelon inventory model with transshipment

In this paper, we propose a continuous-time model for a multi-echelon inventory system. Items stored in the inventory are deteriorating and deteriorated items are screened out continuously from the inventory. Lateral transshipments are allowed among the warehouses in the same echelon. The transshipment rate depends on the inventory level of the corresponding warehouses. In this study, we give some sufficient conditions for the equilibrium points of the system to be stable. By aggregating the warehouses in an echelon, a procedure is developed for finding the equilibrium value of inventory level at each warehouse. The results can be applied in determining the transshipment rate among the warehouses.

On human decision behavior in multi-echelon inventory management

We conducted experiments in a multi-echelon inventory system with two scenarios: no information sharing and full information sharing. The results indicate that the bullwhip effect persists even after traditional causes have been excluded. Moreover, full information sharing cannot guarantee performance improvement. From a behavioral perspective, we analyze different causes of the bullwhip effect for the two scenarios with linear regression models. In addition, we find that the attitudes on two important inventory factors (the on-hand inventory and the stockout) are reversed in the two scenarios. These results may provide managers with useful new ideas for improving performance.