Deconstruction and Recovery Information Modelling (DRIM)

Demolition waste comprises significant proportion of valuable building materials that could be re-used for new constructions and refurbishment. However, no such tool exists that can help in the identification of valuable building materials for the purpose of reuse and recycling. This project developed a BIM-based tool for identifying and reclaiming valuable materials at the end-of-life of buildings.

Project cost

£800,000

Funding bodies

• Innovate UK (£575,000)
• EPSRC (£225,000)

Academic partner

Coventry University

Duration

2016–2018

Industrial partners

• Waste Plan Solutions (WPS)
• Sustainable Direction Limited (SDL)

Project details

The DRIM tool leveraged the advanced machine learning technique to develop a tool that supports the architects, design engineers and other stakeholders at the building design stage in respect of building material choice. The tool also provided support for the Refurbishment and Demolition Surveyor and Engineers at the end-of-life of buildings.

The component of the DRIM tool that supports material analysis and selection at the building design stage was integrated into BIM software environment in the form of an add-in in Revit 2017. This component of the DRIM tool uses deep learning models to predict the amount of materials and their categories (direct reuse, recycle and waste to landfill).

When given basic information about the building such as gross floor area volume, number of level, building type and building usage type, the tool generates the amount of various categories of the recoverable material. The deep learning models are integrated to a web framework and the solution is presented as a web platform.

The DRIM tool solution is currently at the trailing stage. The first set of trailing has been done and the feedback obtained is now being used to further refine the machine learning models to improve accuracy and performance.