A bi-objective robust optimization approach for the management of infectious wastes with demand uncertainty during a pandemic

Abstract

The current global COVID-19 pandemic attracts public attention to the management of waste generated by health-care activities. Due to the hazardous nature, infectious waste requires the design of a multi-tiered system to provide cost-efficient and eco-friendly services of waste collection, transportation, treatment, and final disposal. However, the impact of uncertainties has not been well studied in the existing literature. Considering the presence of random waste generation during a pandemic, we aim to answer the following questions: 1) where to locate temporary transfer stations and temporary treatment centers; 2) how to plan collection tours among the small generation nodes and temporary transfer stations; 3) how to plan the direct transportation from large generation nodes to treatment centers; 4) how to transport waste from temporary transfer stations to treatment centers, and 5) how to transport wastes from treatment centers to disposal facilities. The relevant cost and associated risk are respectively formulated and assessed using a scenario-based bi-objective robust approach. The complexity of the resulting mathematical model motivated the adaption and comparison of three multi-objective optimization approaches, including the goal programming method, a lexicographic weighted Tchebycheff approach, and an augmented ϵ -constraint solution technique. A case study based on the real situation in Wuhan, China, during the COVID-19 outbreak is conducted to demonstrate the workability of the proposed model and provide managerial insights for infectious waste management. The computational results show that our proposed model can more than double the demand fulfillment rate at an approximately 40% lower cost when facing a distinctively high increment in the amount of infectious waste.

Vehicle Routing Optimization for Pandemic Containment: A Systematic Review on Applications and Solution Approaches

“…Uncertainty: Considering the uncertainty in VRP operations during a pandemic, the focus was mainly placed on stochastic customer demand [77,[79][80][81]88,99]. Therefore, fuzzy chance-constrained programming [81], scenario-based approaches [77], and Monte Carlo simulation [99] were utilized to address the severe demand uncertainty. Another stochastic parameter, travel time, was investigated in Shen et al [66], and stochastic programming was used to determine the value of travel time.…”

“…Note that only one work introduced dynamism in the weights, which were dependent on the real-time information of customers' request [85]. Other decomposition-based methods such as Chebyshev method [52], Membership function [82] and Goal Programming(GP) [77,81,105] were also utilized to construct a compromising model for solving multi-objectives problems.…”

“…Regarding waste disposal, seven papers discussed the infectious waste (e.g., medical waste) collection activities during a pandemic scenario. These articles mainly explored various ways to mitigate the transmission risk presented by infectious medical waste, including the use of temporary facilities [74,77,78,81] and outsourcing [105].…”

“…The most common economic factor was cost minimization. The often used cost elements were transportation [93,108], fixed and variable outsourcing [105], the use of vehicles [52,87,105], fuel consumption [109], penalty cost for arriving early or late [76], facility opening costs [77], and shortage costs for relief suppliers [79], etc.…”

“…Notably, three out of fourteen articles that considered health-related factors addressed the problem of waste collection [77,78,105]. A common assumption in these studies is that minimizing the total population exposed during medical waste transportation can reduce the risk of infection, thus improving public health.…”