Residential Waste Dumping Behaviors for Waste Sorting and Relevant Managerial Insights

Waste sorting at sources of individual residential areas is a popular strategy worldwide for improving the sustainability of megacities. However, few effects can be produced if the strategy is implemented by relying solely on propaganda and civic awareness. Along with the strategy, mandatory regulations of ‘fixed waste dumping sites and fixed dumping time windows’ and encouraging bonuses for ‘accumulating points in green accounts’ have been implemented at pilot residential communities in Shanghai in 2018. Impressive effects regarding waste sorting and recycling have been achieved in these communities under such measures. In early stages after policy regulations, the general waste dumping behaviors of whole residents are nearly unknown, which poses great challenges for decision-makers to determine appropriate waste dumping time windows and schedule manpower and material resources for daily waste collection operations. Traditional simulation approaches have difficulties for the problem due to the population scale and the complexity of individual behavior. To cope with the problem, we design an analytical framework for residential waste dumping behaviors based on the internet of things and data-mining technologies. In the data-mining component of the framework, waste dumping behaviors are visually illustrated by the probability distributions of dumping time, dumping time interval, dumping amounts and dumping frequency, the time-serial trends of dumping amounts, and the weekly patterns and holiday patterns of dumping amounts. A real case of a pilot residential community in Minhang district of Shanghai is employed to demonstrate the effectiveness of the proposed analytical framework. Based on the study of residential waste dumping behaviors, relevant managerial insights can be effectively offered for decision-makers and practitioners.