HARNESS

High-quality AI-assisted Reuse for Next-generation Engineering Steel supply

Project Overview

Steel reuse remains limited by manual, fragmented workflows that undermine efficiency and confidence. Fabricators face challenges including manual cataloguing, poor traceability, late-stage design decisions, and ad-hoc off-cut management.

The HARNESS project integrates AI into the fabrication workflow, transforming reuse into a systematic, scalable, and data-driven process. Working within Caunton Engineering’s operational fabrication environment, the project will demonstrate how artificial intelligence can increase reuse rates, shorten lead times, and provide clear economic rules for harvesting high-quality reusable steel.

Improving Key Fabricator Processes

The HARNESS workflow addresses these fabrication challenges that currently hinder steel reuse:

Stock Cataloguing: AI is trained to extract data from images, automatically identifying geometry, defects, and certification details. This replaces manual inspection and Excel-based logging, speeding up cataloguing and improving traceability.
Matching to Projects: Algorithms match reclaimed sections to live IFC/Tekla models. This supports early design-stage decision-making by ranking best-fit components and quantifying modification effort—reducing overdesign and project delays.
Off-cut Management: Currently, short steel lengths are often scrapped by default. HARNESS applies AI heuristics to recommend when off-cuts are viable for reuse and allocates them efficiently across projects.
Design Influence: The system provides feedback loops to designers, enabling layout adjustments and informed design iterations.
Carbon and Cost Impacts: We will benchmark HARNESS workflows against current practice, evaluating embodied carbon savings and economic benefits for fabricators.

Research & Validation

The project will test AI-enabled workflows under real operational conditions, using 350 tonnes of reclaimed steel at Caunton’s facility.

Expected impacts:

Reuse rate improvement: +25–40%
Time savings: 50–70% in stock-to-design matching
Carbon savings: ~630 t CO₂e avoided (350t steel).

Engagement Opportunities

We welcome input from stakeholders across the steel and construction supply chain, including fabricators, designers, main contractors, developers, and industry bodies.

Opportunities to participate include:

Contributing to expert interviews to help assess the commercial value of the proposed solution
Attending an in-person demonstration at Caunton’s facility in Nottingham, showcasing the AI-enabled workflow in operation

Header image by awesomecontent on Freepik