1.1 Supply Chain Management:
Supply chain management is defined as the management of the different flows in a supply chain with a view to improve the long term performance of the individual firms and the supply chain as a whole. A supply chain is essentially a set of three or more companies directly linked by one or more of the upstream or downstream flows of products, services, finances, and information from source to customer. Multi-objective optimization problems arise and the set of optimal compromise solutions (Pareto front) has to be identified by an effective and complete search procedure in order to let the designer to carry out the best choice [19]. Supply chain management involve the planning and execution of all activities involved in sourcing, procurement, production, and logistics. It also includes the crucial components of coordination and collaboration with different channel partners. These partners are suppliers, intermediaries, third-party service providers, and customers. In total, supply chain management integrates and oversees supply and demand within and across companies. Supply Chain Management is primarily concerned with the efficient integration of suppliers, factories, warehouses and stores so that merchandise is produced and distributed in the right quantities, at the right locations and at the right time, so as to minimize total operating cost subject to satisfying service requirements. Integration of supply chain not only reduced the costs, it also accelerated the creation of value for the company, supply chain partners, and all stake holders [11]. Supply chain comprises elements of the business processes, including people, firms, resource, and activities of management and planning. Depending on the features considered for the supply chain, each element has different costs [22]These elements work in synergy in making the product or service from the manufacturer to the final customer with the help of important stack holders like suppliers, logistics partners, distributors and wholesalers [10]
1.2 End-to-End Supply Chain: End-to-End supply chain encompasses an entire integrated process across the supply chain that include procurement of resources, scheduling, production, and delivery of final product to the customer zones. Sometimes, End-to-End supply chain also takes care of after-sales services, reverse logistics, etc., depending on the need of the business. The supply chain entities from the design of the product to sale of the product, and product returns, if any. In optimizing End-to-End supply chain, the above components need to be integrated, and then proceeded with the cost optimization to realize the increased revenue and improved profitability.
1.3 Supply Chain Optimization:
Supply chain optimization is the optimization using mathematical methods to ensure the optimal operation of sourcing, procurement, manufacturing, and distribution. Supply chain operations are optimized to minimize operating costs, minimization of inventories, minimization of operating expenses like inventory, transportation, manufacturing, distribution, etc., and maximize gross margin, maximization of revenue, maximization of ROI. In order to minimize the costs, it is very required to simultaneously optimize several entity costs, including procurement, production, distribution related costs across the supply chain [8]. Supply chain optimization is perceived as integrated approach not a siloed process, and always ensure continuous improvement in cross functional optimizations [3] Supply chain optimization may include refinements at various stages of the entire product lifecycle. Supply chain optimization is a supply chain general problem involving cost reduction and cost control to provide products and services to customers at the lowest price possible for the customers, but the highest profit for the business. Optimizations, in general, involve the usage of appropriate computer software tools with proper parameter settings. Supply chain problems are complex and difficult due to the number of entities in the supply chains, their lead times at each node of the supply chain at both upstream and downstream side, complex inventory management requirements at each entity, stochastic in-demand nature, highly diversified logistic options, etc. Many independent entities in the supply chain each of which tries to maximize their own objective functions or interests in business transactions and many of their interests are conflicting when the entire supply chain is considered. Therefore, for a specific supply chain, giving an optimal design configuration is very demanding. Supply chain cost optimization should enable the supply chain managers to ensure that the supply chain is ‘responsive’, ‘agile’, and ‘reinvented’ for present and future operations [12].
1.4 Non-dominated Sorting Genetic Algorithm (NSGA-II): Non-Dominated Sorting Algorithm-II (NSGA-II) is a developed version of NSGA which features fast and elitist multi-objective non-dominated sorting genetic algorithm without selecting the sharing partners [5]. A Methodology & Algorithm presented by [5] is used to Optimize the Supply chain entities such as Total operating costs, information sharing costs, profit, revenue etc. using NSGA-II. A small initial population size is used in the beginning and the population size is increased where at each step the percentage increase in the number of non-dominated solutions with respect to the previous step is noted. If this percentage increase falls below a certain pre-specified amount, then further increase in population size is not necessary. Srinivas and Deb presented NSGA (Non-dominated Sorting Genetic Algorithm) which was based exactly on MOGA except for a few changes in fitness assignment leaving the rest of MOGA unchanged. NSGA-II is the second version of the famous “Non-dominated Sorting Genetic Algorithm” based on the work of Prof. Kalyanmoy Deb. NSGA-II is a fast and elitist multi-objective evolutionary algorithm. The multi-objective optimization describes a set of solutions that shows the best trade-off between the competing objectives [14]. The most recent implementation of multi-objective genetic algorithm is Non-dominated Sorting Genetic Algorithm-II or NSGA-II that is known as one of the best methods for generating the Pareto-frontier. With the application of non-dominated sorting technique and crowding distance to rank and choosing population front NSGA-II help the decision makers to better understand the relationship among the selected parameters [18]The NSGA-II algorithm ranks the individuals based on dominance. It also calculates the crowding distance for each individual in the new population. Crowding distance gives the GA the ability to distinguish individuals that have the same rank (i.e. those that reside in the same frontier set). With an initial set of random solutions, generally called population, NSGA-II starts. Each individual in the population is called a chromosome. In NSGA-II, the algorithm converges to one solution after a few generations and the chromosomes’ structures in the last generation are similar [9]. Each chromosome consists of genes and as it is known, a gene in a chromosome is characterized by two factors: locus, the position of the gene within the structure of chromosome and allele, the value the gene takes [21].
A brief explanation of the steps involved in NSGA-II is presented below:
- A parent population called Pt is randomly generated and an offspring population Qt is created from it.
- Both populations Pt and Qt and combined into population of size 2N where N is the population size. This new population is called Rt.
- The population Rt undergoes non-dominated sorting where all members are classified and put into fronts.
- The best N individuals from Rt are selected using the crowding tournament selection operator and from the parent population of the next generation Pt+1.
- The steps 1-4 are repeated until the termination criteria have been satisfied.
The source code for NSGA-II is freely available for research purposes at the KanGal (Kanpur Genetic Algorithms Laboratory) website. The website is maintained by Dr. Kalyanmoy Deb and is a portal to his continuing research in GAs. The code is implemented in the C programming language. The motivation for using NSGA-II in this article is because it’s performance has been tested on several test functions and has given accurate results in generating the Pareto front as can be seen in Deb et al, 2002 and it is well supported by literature having been used in many real-world applications. However, the decision-maker can adopt a different plan from the efficient Pareto-optima set by changing the objective functions’ weight [2].
1.5 Supply Chain Operations – Revenue generation:
Supply chain operations are perceived as revenue generation in business operations [4]. Various objectives like minimization of total operating costs, minimization of information sharing costs, maximization of revenue, etc. form the necessary pre-requisites for effective and efficient strategies in supply chain practices. But these are conflicting objectives if total efficiency and effectiveness of supply chain performance are to be achieved. Adhering to one strategy does not ensure the perceived/improved performance of the supply chain. Therefore, conflicting objectives need to be considered set-wise so as to arrive at complete and effective SCM practices. The key objective of supply chain optimization is to maximize profitability. It is not optimizing any single entity of the supply chain but rather all the entities taking part in the supply chain. This is possible when all the entities in the supply chain are optimized for the expected performance to reap profits, rather than individual entities. Lower operational and material costs in supply chain operations invariably increase the enterprise’s profit margins and financial performance [17]. Activities, elements, and procedures in the supply chain once optimized, shall result in firm performance and efficiency, which in turn, refer to revenue generation, and profitability [15]. There must be a complete integration among the entities so that information of the supply chain transactions can be shared in real-time always in order to meet the highly fluctuating demand of the customers.
Achieving objective of integration, commitment and coordination of other supply chain partner companies are very important [20]. Therefore, to maximize the profitability of the supply chain it is not to optimize these individual drivers separately, but as a whole. Objective functions catching these drivers will also have to be optimized simultaneously.