Predictive Analytics for Quantitative Trade-in-to-Upgrade Decision Making in Intelligent Disassembly-to-Order Systems
Loading...
Authors
Tozanli, Ozden
Issue Date
2019-12-09
Type
Thesis
Language
en_US
Keywords
Disassembly-to-order systems , Discrete event simulation , End-of-life product recovery , Sustainable supply chains , Trade-in policymaking
Alternative Title
Abstract
The accelerated growth of technological advancements has triggered the expansion of customer demand leading to highly complex supply chain networks. One viable way original equipment manufacturers (OEMs) can respond to changing purchasing habits is to redesign their strategic and operational activities to build far-reaching information and resource avenues allied with effective marketing policies. These newly implemented policies need to comply with extended producer responsibility (EPR) guidelines that also well align with rising consumer awareness towards green consumption. To achieve this, manufacturers must create efficient end-of-life product (EOLP) return structures and ensure value creation through product recovery operations to dwindle the cascading waste of discarded products. From an environmental viewpoint, retrieving the value embedded in returned items through remanufacturing or recycling has been proven to be effective in reducing the amount of industrial solid waste. EOLP processing operations are heavily reliant on customers' participation in returning outdated devices making product collection a crucial step in point-to-point supply chains. To entice end-users, the OEMs need to design environmentally and economically benign product take-back strategies that would spark the volume of product returns. These constraints dictate two structural challenges: how manufacturers and consumers can become active participants of EOLP treatment activities, and how fast and efficiently OEMs can respond to the changing market and capital needs while preserving their sustainability levels. In terms of active participation, trade-in incentives can help stimulate additional revenue channels for OEMs through product remanufacturing while helping companies comply with the EPR legislations. Trade-in policies are set forth as part of long-term marketing strategies and include incentive programs that aim at enticing current and potential customers to trade-in their used products with newer generations at a discounted price or for instant credit. Within the context of purchasing behavior, trade-in programs positively impact customers' buying decisions by granting buyers the ability to claim the scrap value of their existing devices. Particularly in oversaturated industries such as electronics and automotive, take-back incentives are a pipeline for OEMs to generate significant residual value by reselling remanufactured products on secondary markets. Moreover, offering special discounts or credits in lieu of old devices fuels new product sales by creating an additional revenue stream. Still, in today’s fast-changing market dynamics, inept trade-in practices that fail to eliminate the ambiguity surrounding the prediction of the true quality of returned products bring functional and financial burdens to organizations. The conventional intransigent trade-in schemes fail to address this uncertainty leading to a number of unnecessary inspection, disassembly, and shipment steps resulting in increasing complexity and product recovery cost. Achieving an accurate trade-in scheme is a highly complex multi-dimensional problem requiring novel solutions that traditional manufacturing and supply chain technologies are incapable of offering by design. Such challenging task inevitably necessitates strategic initiatives that stem from the utilization of cutting-edge groundbreaking information technologies for rapid response to customer needs and reduced complexity across all operational layers. Despite the numerous methodologies investigating the potential value gain from remanufacturing and product acquisition pricing policies, there is no study in related literature that incorporates trade-in programs into an intelligent remanufacturing structure. A majority of previous studies propose preventive models with pre-determined and rule-based explicit model parameters hindering the practicability of the substantial volume of data generated by the increased use of technological tools. These models, inevitably, fall short in successfully incorporating long-term manufacturing goals into sustainable business strategies. With this motivation, the architectural framework this dissertation introduces addresses a predictive product recovery model for product returns to enable an autonomous, sensor-embedded, and decentralized disassembly and remanufacturing system. The main objective of this research is to investigate the feasibility of cost- and resource-effective end-of-life product management systems in a smart reverse logistics network where trade-in rebate decisions take place in an autonomous ecosystem. This research, while filling the emerging gap in the utilization of current digital technologies to determine quality-dependent acquisition strategies, also provides a novel quantitative analysis on the efficiency of trade-in policymaking. This model can be employed in manufacturing industries for precise assessment of value creation amid digital advancements in a future-oriented platform. Due to its highly saturated formation, the consumer electronics industry offers a more suitable platform for this study. Therefore, this study examines a trade-in model for a specific technological product, game console, with the help of a case study. First phase of the dissertation evaluates the performance degradation pattern of discarded electronics products in a ubiquitous manner through timestamp data enablers. To handle this highly complex large-volume data, a discrete-event simulation model is developed from the original equipment manufacturer viewpoint. The model aims to examine the behavior of returned devices as well as the expected overall cost of product recovery operations. Following this, a design of experiments study is utilized for the experimentation using Taguchi’s Orthogonal Arrays (OAs). Employing the findings obtained in the first phase, the second phase of the study deals with trade-in policymaking to determine an engaging quotation for varying quality of returned products from the perspectives of all parties involved in the transaction. To achieve this, an initial model for trade-in-to-upgrade incentives is established for discrete sets of quality standards in case where returned products are grouped into three quality classes based on their usage time. The model is then expanded to compare two product acquisition strategies, namely, trade-in-to-upgrade incentives and instant credits. To achieve a realistic strategy, two rebate models are constructed in a simulation-based game setting to mimic the customer behavior and to obtain the resulting payoffs for the OEM in a dynamic ecosystem. To handle the uncertainty in the customer's decision towards the incentive offer, logistic regression analysis is conducted to maximize the likelihood of the acceptance rate. Finally, trade-in policies are compared to obtain favorable strategies augment revenue streams.
Description
Citation
O.Tozanli, "Predictive Analytics for Quantitative Trade-in-to-Upgrade Decision Making in Intelligent Disassembly-to-Order Systems", Ph.D. dissertation, Dept. of Engineering, Univ. of Bridgeport, Bridgeport, CT, 2019.