Reverse Logistics and Closed-Loop Supply Chain

In the last years, Reverse Logistics and Closed-Loop Supply Chain issues have attracted attention among public opinion, academia and industry. The focus on Reverse Logistics and Closed-Loop Supply Chain originated from public awareness, then faced by governmental legislation with the aim of forcing producers to take-back and manage their EoL products, e.g. Directive 2002/96/EC (now Directive 2012/19/EU) on Waste Electrical & Electronic Equipment (WEEE), and Directive 2000/53/EC on end- of-life Vehicles (ELV). Now, in many industrial sectors, Reverse Logistics and Closed-Loop Supply Chain are considered an opportunity for supply chain cost minimisation (Guide and Van Wassenhove, 2009).

Traditionally, a supply chain is understood in its “forward” form, which corresponds to “a combination of processes to fulfil customers’ requests and includes all possible entities like suppliers, manufacturers, transporters, warehouses, retailers, and customers themselves.” (Chopra and Meindl, 2010).

Reverse Logistics is defined by Tibben‐Lembke (2002) as “the process of planning, implementing, and controlling the efficient, cost effective flow of raw materials, in-process inventory, finished goods and related information from the point of consumption to the point of origin for the purpose of recapturing value or proper disposal”. The integration of Forward and Reverse supply chains, simultaneously, results in the creation of a Closed-Loop Supply Chain. In Guide et al. (2003) the Closed-Loop Supply Chain Management is defined as “the design, control, and operation of a system to maximise value creation over the entire life cycle of a product with dynamic recovery of value from different types and volumes of returns over time”.

According to Govindan et al. (2014b), the contributions from the literature to the issues of Reverse Logistics and Closed-Loop Supply Chain can be classified as follows:

 Network Designing and Planning: the aim of designing is to determine strategic decision variables, such as facility location and facility capacity. In the planning stage, the most important decision variables are the quantities of flows between supply-chain network entities known as mid-term decision variables. Some studies regard designing and planning stages simultaneously, and some concentrate on one of them in depth. The topic of Supply Chain Network Design is deepened in 3.3.4.

 Network Planning: it is a sub-category of the previous one, in which the planning level decisions, such as quantity of flows between network entities, are studied without regarding any strategic or operational decisions.

 Production planning and Inventory Management: management issues such as finding reorder point, base stock, and economic order quantity without regarding production-planning subjects.

 Price and Coordination: this category includes studies that focus on the determination of the price of products and the coordination of win-win strategies to balance profit margins between two entities of a supply chain network (e.g. a remanufacturer and a retailer of second market). Usually, in such problems, optimum price and coordination strategies are determined.

 Decision-making and performance evaluation: this category includes the research on the evaluation of the performance of various networks and recovery strategies in Closed-Loop Supply Chain.

Depending on the specific topic, different tools, techniques and methodologies are adopted in literature for the design, planning, optimisation and control of Reverse Logistics operations and Closed-Loop Supply Chain Management.

3.3.3.1 Tools/Techniques/Methodologies in Reverse Logistics and Closed-Loop Supply Chain Both Reverse Logistics and Closed-Loop Supply Chain have been faced by many authors in recent years. Srivastava (2007) presents an extended review of these issues. The author focuses on studies related to the mathematical modelling for network design and planning problems. The author builds a taxonomy based on mathematical tools/techniques. Results show that the methodologies applied the most in this context are the following: Mixed Integer Linear Programming, simulation, sensitivity analysis, algebraic equations, heuristics and meta-heuristics, dynamic programming, Markov chains, and game theory. Although to a lesser degree, also Petri-net, Analytic Hierarchy Process, Fuzzy reasoning, and neuro-fuzzy are used. Classifying them on the basis of the decision level they deal with, it is possible to notice that 35%, 33% and 32% is the portion of studies having operational, tactical and strategic focus, respectively.

3.3.3.1.1 Linear Programming

Govindan et al. (2014b) analyse studies published between 2007 and 2013 on Reverse Logistics and Closed-Loop Supply Chain. According to the survey, 18.8% of papers deal with the Design and Planning of Closed-Loop Supply Chains, and the 69.4 % of these researches are based on linear modelling, such that it is possible to claim that the Linear Programming approach can be introduced as the dominating modelling approach for the design and planning problems of Reverse Logistics and Closed-Loop Supply Chain.

3.3.3.1.2 Exact solutions VS heuristics and meta-heuristics

The authors propose a further classification of methodologies, according to which methods leading to extract solutions and heuristics and meta-heuristics are split in two categories. The survey shows that in case of large complex problem, utilising heuristic and meta-heuristic algorithms is unavoidable, but these methods do not ensure knowledge about the quality of the found solutions. Despite the fact that analytical and exact methods are rarely applicable to real-sized instances of a problem, they are still largely studied and proposed in literature (41.6% against 11.2% of heuristics and meta-heuristics).

3.3.3.1.3 Single VS Multi period, product and objective

A further classification can be made based on the number of periods, products and objectives considered in the problem modelling. Govindan et al. (2014b) classified recent papers on the basis of single- and Multi-objective models, for Single and multi-period, and for single- and multi-product problems. The trend in recent literature is shown in Figure 3.3.1, Figure 3.3.2 and Figure 3.3.3, in which the incidence of each approach is measured by the number of papers per period.

Figure 3.3.1 - Trend of Single- and Multi-Period problem modelling (from Govindan et al. (2014b))

Figure 3.3.2 - Trend of Single- and Multi-Product problem modelling (from Govindan et al. (2014b))

There is a balance between Single- and Multi-Period problems. That proves the equilibrium in the ratio of strategic and planning models: the former are characterised by single-period problems, the latter by multi-period modelling. However, a negative trend for Single-period approaches has been recently noticed, which demonstrates that dynamic approaches are more representative of the reality. Finally, the majority of recent papers present single-product models (65.4%) and only few studies consider multi-part products (just 5.4%). This result is probably caused by the computational difficulties that Multi-

Figure 3.3.3 - Trend of Single- and Multi-Objective modelling (from Govindan et al. (2014b))

3.3.3.1.4 Reverse Logistics, Closed Loop Supply Chain and Multi-Objective problems

Multi-objective approaches are still a minor part in recent publications: 87.6% of papers deal with Single- objective approaches while only 12.4% present Multi-objective tools. These numbers demonstrate that Multi-objective decision-making is still a gap in literature (Govindan et al., 2014; Kumar and Kumar, 2013). Real world problems are rarely single objective, therefore implementing Multi-objective functions instead of single objective ones is a priority in research. The approaches for dealing with Multi-objective problems and achieving the optimal solutions (e.g. Pareto optimal solutions) need to be revised to produce more robust and applicable methods.

One of the most interesting extension in objective functions regards the introduction of sustainable and environmental objectives. According to Govindan et al. (2014b) “it is expected that researchers regard appropriate environmental, social, and green-based objectives in their analyses, which can be a critical future avenue for all entities in the Reverse Logistics and Closed-Loop Supply Chain network”, who concludes the paper with “the integration of different levels of decision-making and defining new decision variables are future opportunities for the decision variables category. Paying attention to multi objective problems, utilising new approaches, and applying more green, sustainable, and environmental objectives can be the future directions in single and multiple objective problems”.

Among the potential topics of Reverse Logistics and Closed-Loop Supply Chain, this research focuses on the issue of Network Design and Planning, which is discussed in the following section.