A Memetic Algorithm for Integrated Production Distribution Problem in a Supply Chain

Supply chain is a complex system in which most of the activities are inter-related, and changes in one of these activities can affect the performance of the other processes. Thus, integrated management strategies in a supply chain can yield considerable advantages throughout the system as supply chain members and customers become more integrated. In this study, a memetic algorithm is proposed to solve the integrated production-distribution problem. The objective of the problem is to find optimal production quantity, customer delivery quantity, and schedule to minimize the total system cost, which is composed of production setup cost and variable production cost, inventory holding costs, and distribution cost. The effectiveness of the proposed algorithm is tested on the existing data sets. According to test results, the proposed algorithm is a very effective method to solve integrated production-distribution problems. To assess to benefits and applicability of the method on the real-life problems, a case study is conducted in a Turkish water manufacturing company.

Designing Valid Humanitarian Logistics Scenario Sets

Literature about humanitarian logistics (HL) has developed a lot of innovative decision support systems during the last decades to support decisions such as location, routing, supply, or inventory management. Most of those contributions are based on quantitative models but, generally, are not used by practitioners who are not confident with. This can be explained by the fact that scenarios and datasets used to design and validate those HL models are often too simple compared to the real situations. In this chapter, a scenario-based approach based on a five-step methodology has been developed to bridge this gap by designing a set of valid scenarios able to assess disaster needs in regions subject to recurrent disasters. The contribution, usable by both scholars and practitioners, demonstrates that defining such valid scenario sets is possible for recurrent disasters. Finally, the proposal is validated on a concrete application case based on Peruvian recurrent flood and earthquake disasters.

Demand Forecasting in Supply Chain Management Using Different Deep Learning Methods

Supply chain management (SCM) is a fast growing and largely studied field of research. Forecasting of the required materials and parts is an important task in companies and can have a significant impact on the total cost. To have a reliable forecast, some advanced methods such as deep learning techniques are helpful. The main goal of this chapter is to forecast the unit sales of thousands of items sold at different chain stores located in Ecuador with holistic techniques. Three deep learning approaches including artificial neural network (ANN), convolutional neural network (CNN), and long short-term memory (LSTM) are adopted here for predictions from the Corporación Favorita grocery sales forecasting dataset collected from Kaggle website. Finally, the performances of the applied models are evaluated and compared. The results show that LSTM network tends to outperform the other two approaches in terms of performance. All experiments are conducted using Python's deep learning library and Keras and Tensorflow packages.

A Cluster First-Route Second Solution Approach for the Multi-Period Home Healthcare Routing and Scheduling Problem

In the home healthcare routing and scheduling problem (HHCRSP), nurses are allocated to a variety of services demanded by clients during a planning horizon. The properties of this problem resemble vehicle routing and nurse scheduling. To propose an efficient solution, the authors consider various issues such as multi-depot, travelling time, time windows, synchronisation, the qualification levels, and other features of nurses and clients. In addition, the continuity of care and work overload should not be ignored in this perspective. First, the authors developed a model in which the continuity of care is redefined by considering connected (synchronous) jobs and the work overload is formulated considering nurse-to-patient staffing ratio. Second, a two-stage solution approach based on a cluster-assign algorithm and variable neighbourhood search (VNS) and variable neighbourhood descent (VND) algorithms are tested on a series of large-scale instances. Computational results present the relations and trade-offs among the aforementioned issues.

Demands and Sales Forecasting for Retailers by Analyzing Google Trends and Historical Data

A supply chain includes several elements such as suppliers, manufacturers, retails, and customers. Forecasting the demands and sales is a challenging task in supply chain management (SCM). The main goal of this research is to create forecasting models for retailers by using artificial neural network (ANN) and to enable them to make accurate business decisions by visualizing future data. Two forecasting models are investigated in this research. One is a sales model that predicts future sales, and the second one is a demand model that predicts future demands. To achieve the mentioned goal, CNN-LSTM model is used for both sales and demand predictions. Based on the obtained results, this hybrid model can learn from very long range of historical data and can predict the future efficiently.

Exploring a Downstream Demand Inference Strategy in a Decentralized Two-Level Supply Chain

A coordination approach for forecast operations, known as downstream demand inference, enables an upstream actor to infer the demand information at his formal downstream actor without the need for information sharing. This approach was validated if the downstream actor uses the simple moving average (SMA) forecasting method. To answer an investigative question through other forecasting methods, the authors use the weighted moving average (WMA) method, whose weights are determined in this work thanks to the Newton's optimization of the upstream average inventory level. Starting from a two-level supply chain, the simulation results confirm the ability of the approach to reduce the mean squared error and the average inventory level, compared to a decentralized approach. However, the bullwhip effect is only improved after a certain threshold of the parameter of the forecasting method. Still within the framework of the investigation, they carry out a comparison study between the adoption of the SMA method and the WMA method. Finally, they generalize their results for a multi-level supply chain.

Forecasting Sales and Return Products for Retail Corporations and Bridging Among Them

The purpose of this study is to show how we can bridge sales and return forecasts for every product of a retail store by using the best model among several forecasting models. Managers can utilize this information to improve customer's satisfaction, inventory management, or re-define policy for after sales support for specific products. The authors investigate multi-product sales and return forecasting by choosing the best forecasting model. To this aim, some machine learning algorithms including ARIMA, Holt-Winters, STLF, bagged model, Timetk, and Prophet are utilized. For every product, the best forecasting model is chosen after comparing these models to generate sales and return forecasts. This information is used to classify every product as “profitable,” “risky,” and “neutral,” The experiment has shown that 3% of the total products have been identified as “risky” items for the future. Managers can utilize this information to make some crucial decisions.

Logistics Providers in Syria Humanitarian Operations

Logistics providers have become an important element in completing humanitarian relief work in countries experiencing armed conflict. Delivery aid assistances need to build logistics capacity and critical supply chain functions that help to meet the unconfirmed requirements of beneficiaries at right place, on right date, and with right fees. To reach the research goal, the authors will determine the weights of customer requirements (CRs) using the DEMATEL method, which considers the influences of inconformity and the causal relationship between CRs. This chapter employs quality function deployment (QFD) to integrate the voice of CRs and supplier criteria TRs using house of quality charts. This chapter focuses on case of humanitarian organizations collaborate with logistics service providers (LSPs) to maintain and enhance their performance by identify the crucial factors that effect on LSPs selection and their specified from the perspective of humanitarian relief organizations activated in Syrian humanitarian operation.

Mixed Delivery and Pickup Vehicle Routing Problem With Limited Flow and Assignment of Drones in an Urban Network

A new variant of the delivery and pickup transportation problem called mixed delivery and pickup routing problem with unmanned aerial vehicles in case of limited flow is introduced. The objective is to minimize operational costs including total transportation costs and service time at each point. This variant is a solution for the urban congestion, and consequently, it is an improvement of the general transport system. First, the problem is formulated mathematically. It is considered as NP-hard; therefore, the authors proposed an iterated local search algorithm to solve the problem of mixed pickup and delivery without drone. Then, a vehicle first-drone second algorithm is used to solve the mixed delivery and pickup problem with drone. The performance of the method is compared through numerical experiments based on instance derived from the literature as well as on a set of randomly generated instances. Numerical results have shown that proposed metaheuristic method performs consistently well in terms of both the quality of the solution and the computational time when using drone with vehicle.