Digital twin-enabled urban construction and demolition waster trading

Abstract

The proliferation of new construction, renovation, and demolition projects has led to a year-on-year increase in construction and demolition waste (CDW) generation, imposing significant pressures on environmental governance. Current CDW management remains in a low-digitalization, low-efficiency phase, with limited adoption of digital technologies. Information silos among stakeholders in waste disposal pose substantial challenges to supervision, information sharing, and recycling practices. Beyond low digitization levels, the insufficient recognition of CDW value has constrained its circulation. Currently, advancements in recycling and remanufacturing technologies enable CDW to serve as a primary raw material for producing high-quality, diverse recycled products. Further, the growing emphasis on the circular economy and rising demand for recycled materials also provide a robust foundation for CDW market circulation. Against this backdrop, this thesis conducts a systematic exploration of digital twin–based CDW trading through a structured approach comprising problem identification, framework development, theoretical analysis, and demonstrative case. Firstly, a digital twin–enabled CDW trading framework and related workflows are proposed, leveraging the virtual-real interaction and real-time mapping capabilities to track CDW status via data and model layers, thereby addressing information silos between waste generators and recyclers. Secondly, a reverse supply chain for CDW trading is introduced and analyzed through three dimensions: logistics networks, capital flows, and information-sharing mechanisms. A cost-benefit analysis is conducted to evaluate how internal, external, and market factors influence the waste-generating units. To address cross-regional disposal challenges, the thesis integrates the government’s big data insights on waste generation/disposal capacity with market-driven supply-demand matching to propose a hybrid allocation scheme combining government-led coordination and market-oriented trading. Finally, a digital twin–based CDW trading system is developed using Internet of Things, big data, robotics, and 3D printing technologies. The primary contributions of this study are as follows. First, it is among the first contributions to the market-oriented circulation of CDW, providing a novel mode for CDW management. Second, the research develops a digital twin framework for the full lifecycle management of CDW trading, leveraging digital twin technology to advance the digitalization level of CDW management. Third, this research undertakes theoretical investigations into reverse supply chain design and cross-regional collaborative disposal, addressing low digitalization and stakeholder collaboration gaps via a digital twin framework. Additionally, a Dual Supervised Fine-tuning Proximal Policy Optimization (Dual-SFT-PPO) method is introduced to enhance the training efficiency and policy superiority of the PPO. Experimental results demonstrate that the reward of the output policy achieves an improvement of over 88% compared to the traditional PPO algorithm. Managerial implications for reverse supply chain development, CDW trading market formation, and CDW cross-regional disposal are summarized. Finally, in the demonstrative case, a CDW trading system is developed employing digital twin technology to visualize critical stages in CDW treatment processes, such as CDW generation, trading, supervision, and logistics. At the model level, the latency of the digital twin model is approximately 12 seconds. At the data level, managers can receive real-time error notifications within 5 seconds. Three dedicated APPs are developed for demolition workers, logistics providers, and recyclers.