A customer-centric framework towards robust e-waste management

• Main Author: Guo, Rui
• Other Authors: -
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Industrial engineering::Supply chain
• Online Access: https://hdl.handle.net/10356/158349
• Institution: Nanyang Technological University

Due to the rapid economic growth and technology advancement, the worldwide consumption of electrical and electronic products has been increasing dramatically. Additionally, owing to the shortened life-span of current e-products and accelerated replacement for emerging new functions, the waste electrical and electronic equipment (WEEE) becomes one of the fastest growing wastes globally in recent decades and persists in increasing significantly for long. Awareness of the quick population expansion and electronics product sales, the total volume of WEEE generation will be enormous. To escape the dilemma, the European directive on WEEE management of 2012 has established new collection frameworks to boost the reverse supply chain system efficiency and thus recover more resources out of WEEE. Nevertheless, there are still problems of inefficiency and inadequacy existing in the WEEE collection system in response to complicate customers’ behavior. To address this research gap, two main challenges need to be tackled and a proactive and nonintrusive approach to forecasting the WEEE generation and collection amount and a sustainable circular system construction is proposed in this study. Prior to build a WEEE management system, an accurate forecasting model is required. Along the line of WEEE recovery, customers are the ultimate decision-makers to determine the destinies of the end-of-life (EOL) electronic products. However, inadequacies exist and those are WEEE collection and generation predictions considering uncertain customers’ behavior and deepening discussion of the influence discrepancies of customers’ behavior between an advanced and a fledging recovery system. To bridge these research gaps, this paper aims to propose a customer-centric approach to forecasting the WEEE collection and generation amounts in Taiwan (i.e., a developed recovery system) and Vietnam (i.e., an developing recovery system), respectively, and analysing the influence disparities from these two societies. First, the combination of grey relational analysis (GRA) and principal component analysis (PCA) is employed to deal with factors reflecting customers’ behavior. Second, the forecast model, the grey neural network (NN) optimized by the particle swarm optimiza- tion (PSO) and genetic algorithm (GA), is developed to estimate the collection and generation amounts of waste home appliances during 2021-2030 in Taiwan and Vietnam. The collection amount in Taiwan is predicted to increase 2-fold until 2030 and the generation amount in Vietnam increases 6-fold. Moreover, different types of waste home appliances show distinct growth trends. The framework shows the capability to cope with nonlinear and uncertain characteristics and realize high accuracy. Based on the predictions, this thesis quantifies the potential sustainability improvement given that realization of a well- established recovery system and a high recycling rate. The total revenue from recovery is estimated to reach 300 million in Taiwan by 2030, and 1.2 billion in Vietnam. Additionally, by 2030 the 〖CO〗_2 emissions in Taiwan are expected to reduce 0.7 million tons and in Vietnam will be reduction of 3.5 million tons. Socially, around 400-500 job vacancies can be created in Taiwan and more than 7000 jobs created in Vietnam in 2030 to achieve effective WEEE collection. The result facilitates the decision-making of the involved industries and regulators for better WEEE management and infrastructure investment. Besides, to have the collection trend in an advanced system (e.g., Taiwan) as a reference, it brightens the way to build and improve a formal collection system in a developing country or region (e.g., Vietnam). A sustainable IoT-based closed-loop supply chain (CLSC) system is then established on the basis of the predicted WEEE collection and generation trend. Due to the reality constraints, a simulation model is constructed to consider the customers’ behavior in the IoT-enable system. Different attributes of customers and external factors are discussed to facilitate improving the efficiency of the system. In the system, the quality assessment of WEEE is also explored to provide a standard. To assess the efficiency of the proposed system, a benchmark model named as scenario one is simulated. The proposed model is simulated as scenario two. The economic, environment and social dimensions are assessed to achieve sustainability. The influential degrees of different customers’ attributes to the proposed network are studied. The second-hand price and buy back price are two highly important factors, followed by peer pressure and social influence, which suggests more public advertisement and education campaign. Convenience of the collection facilities is also an important factor. Hence, the curbside bins are suggested to cover more customers. In addition, timely emptying the curbside bins encourages more WEEE returns to formal recyclers, which manifests the advantages of the proposed framework. With comparison to the extant system using life cycle assessment, it is discovered that the IoT-enabled system is more efficient and sustainable from economic, environmental and social perspectives. Appropriate WEEE management using the IoT-enabled system makes more profit and can reduce more greenhouse gas emissions and heavy metal contamination. It is noticed that the economic savings in optimized case of scenario two can reach about 1.6 times in comparison with scenario one. It demonstrates that the energy savings in the best scenario can achieve 7.8 times higher than that of scenario one (the benchmark scenario). And the total 〖CO〗_2 emission in the optimized scenario two is reduced around 97% compared to the benchmark scenario one. Although the total 〖CO〗_2 emission in the optimized IoT-based experiment is more than that in the un-optimized experiment in scenario two, the global warming potential reduction makes up the larger part. In addition, with applying the proposed system, the number of people being protected from diseases caused by heavy metals can be significantly reduced.