About This Project

Demolition in construction remains a significant challenge, particularly in managing material waste. I believe that combining in-situ structural testing with AI can accurately identify structural steel suitable for reuse before demolition occurs. This project aims to develop a prototype tool to test this hypothesis, integrating it into circular economy principles and structural design workflows. The goal is to support scalable circular construction practices and reduce embodied carbon.

Ask the Scientists

What is the context of this research?

Steel is one of the most used construction materials, yet tons are discarded during demolition even when still structurally reusable. Buildings account for 34% of global energy use and 37% of related greenhouse gas emissions (UNEP, 2022, https://www.unep.org/). A growing share is embodied carbon: from material extraction, manufacture, construction, maintenance, demolition, and waste processing (UK Green Building Council, 2021, https://www.ukgbc.org/; RICS, 2017, https://www.rics.org/). While recycling is common, reuse saves more energy, cuts emissions, and reduces waste. Currently, no simple method exists to identify reusable steel before demolition, and post-demolition testing is often costly, sometimes exceeding the price of new steel. This project hypothesizes that combining in-situ mechanical testing with AI analysis can accurately and efficiently identify reusable steel, improving decisions and reducing errors compared to traditional methods.

What is the significance of this project?

Steel is a cornerstone of the UK construction sector, contributing £2 billion annually and supporting over 33,000 jobs (UK Steel, 2023, makeuk.org). While 91% of demolition steel in the UK is recycled, most scrap is exported rather than reused locally, missing opportunities for job growth and carbon reduction. Globally, steel production accounts for about 9% of CO2 emissions, with UK steelmaking responsible for 12 million tonnes in 2019 alone (World Steel Association, 2020, worldsteel.org). Global steel demand is expected to reach 2.5 billion tonnes by 2050 (IEA, 2021, iea.org). The IEA estimates that improving material efficiency, including reuse, could contribute significantly to emissions reductions in the steel sector, potentially accounting for around 30% of combined emissions cuts across key materials by 2060 (IEA, 2021, iea.org). This project aims to support circular economy goals by enabling accurate identification of reusable steel.

What are the goals of the project?

The project aims to develop a prototype digital tool for identifying reusable structural steel prior to demolition. Mechanical data will be collected through in-situ tests measuring strength, hardness, corrosion and structural context information. Laboratory testing on steel samples will be conducted to validate these measurements. An AI model will be developed using machine learning algorithms, trained with positive and negative controls to accurately classify steel reuse potential. The model’s predictions will be validated against laboratory results and expert assessments from structural engineers and sustainability consultants. The prototype will be designed to integrate with digital workflows, allowing engineers and contractors to input data and receive guidance on reusable steel elements, thereby supporting circular construction practices and waste reduction.