AI-enabled Design Optimization for Waste Efficiency (AI-DoWEP)

点击翻译按钮获取中文摘要

137.8 million tonnes of waste are generated from construction, demolition, and excavation (CD &E) activities and account for 62% of the total waste generated in the UK in 2018 (Department for Environment Food & Rural Affairs, 2022). The impact of waste on the construction industry is enormous with an economic cost of about £11 billion per annum, adverse effects on profit, and negative environmental and health impacts (Green Construction Board, 2020). The Waste (England and Wales) Regulations 2011 advocates for a hierarchical approach in waste management -- reduction, reuse, recycling, recovery, and disposal -- with waste reduction being the most preferable option. Consequently, material efficiency is the most preferable solution to effective waste management on construction projects. _This involves using fewer materials, and the adoption of alternative means of design or construction that results in lower material usage and lower wastage level_ (BREAAM, Mat 06).Despite material efficiency being the most preferable option for preventing and reducing waste in the construction industry, disposal of waste is prevalent. Lots of avoidable wastes are generated during the construction stage due to a lack of designing out waste at an early stage (Green Construction Board, 2020). Construction businesses employ manual approaches which are time-consuming, susceptible to errors and not optimal. Designing out waste is a complex optimization problem of selecting optimal dimensional coordination of design and standardization of materials that would generate the least waste.Advances in the field of Artificial intelligence, and big data analytics offer opportunities for optimizing design to reduce waste generation which the proposed project seeks to implement. The project aims to create an AI-enabled system (AI-DOWEP), as a BIM-based solution, consisting of two key elements as follows:a. AI-enabled Design Diagnostic Platform for Waste Efficiency: This diagnoses building designs to estimate its waste efficiency index, and automatically highlight areas for potential waste minimisation. This would be driven by real-time data analytics of available and readily accessible material standards, specifications, shapes and other anthropometrics.b. AI-enabled Design Optimisation Platform: This would leverage AI Evolutionary algorithms to optimise building designs for waste efficiency in line with intended functionalities and design anthropometrics. It would solve the optimization problem of 'what is the optimal design dimensional coordination and material standardization that would generate the least waste'.