DRIFT: Disaster Reconnaissance and Intelligent Forecasting Technology

DRIFT is an international project aiming to deliver an innovative technology‑based service that will provide a more robust and faster process for the seismic assessment of buildings.

The service is an AI‑based system that interprets the vulnerability of structures by combining a cutting‑edge 2D and 3D imagery‑extraction algorithm with automated metadata generation for damage classification and waste optimisation.

DRIFT will deliver an innovative project aiming to help society to be better prepared for, respond to, and recover from disasters.

Post‑disaster assessment of buildings that have been affected by a disaster is crucial in guiding rapid recovery, as well as the optimal utilisation of waste material. DRIFT will deliver a practical and innovative service to enable critical rapid assessment. The service will have advanced capabilities and a unique method of both image and data acquisition, replacing standard and currently outdated, time‑consuming approaches by enabling the collection of reliable and rapid information on structural condition.

DRIFT will become a reference service for “Risk Evaluation”, with a unique repository of collected data. It will also help communities to prepare for and mitigate the impact of natural disasters. The output will lead to improved resilience, as well as efficient recovery and reconstruction.

DRIFT stands out as a unique solution in the 3D scanning and structural assessment sector due to its focus on rapid, automated post‑disaster analysis.

Unlike conventional 3D scanning tools, which prioritise high‑precision data capture for construction and surveying, DRIFT is specifically designed to deliver fast and reliable assessments of structures following earthquakes and other disasters. Traditional solutions typically require significant time, specialised equipment, and expert operators, making them less practical in emergency contexts. DRIFT addresses these limitations by generating datasets with a substantially lower computational burden than alternative algorithms and at a significantly reduced cost.

Project consortium

The DRIFT consortium comprises three Eureka countries: South Korea, Turkey, and the UK. The UK project will be led and coordinated by Aralia Systems (UK). The consortium is an international team with strong interdisciplinary expertise, bringing together academic researchers, industry specialists, and R&D departments. The geographical distribution of the partners adds significant value to the research, as well as substantial opportunities for impact. The international dimension will also allow the identification of limitations, challenges, and strengths that are characteristic of local communities and regions.

Project objectives

1. To extend the capabilities of an AI-based high-performance 3D algorithm, developed by the UK leading partners, by enabling to capture structural and safety data of buildings.
2. Establish a robust digital framework for recording damage in buildings after a seismic event with advanced geographic information for the vital emergency phases.
3. To develop a multi-criteria automated procedure for seismic assessment of buildings by integrating remote sensing and on-site surveying.
4. To define a harmonized and advanced database of seismic risk by combining open-source records and in-house generated case-studies.
5. To rapidly assess and provide solutions to transform demolition waste into optimal retrofitting materials.

Methodology

DRIFT leverages neural networks for real‑time, automated structural analysis, overcoming the main limitations of existing scanning techniques and enabling quicker data‑sharing and coordination efforts. In particular, DRIFT presents an innovative approach that provides a complete, robust, and fast processing pipeline capable of accommodating both large‑scale models and high‑resolution surface detail. It uses raw 2D images captured by a smartphone to extract structural metadata in close to real time. QGIS has been selected as the platform environment owing to its thriving open‑source community built on cooperation and knowledge‑sharing.

The consortium will investigate the use of recycled aggregates derived from construction waste, as this represents the largest component of debris and waste. This approach will offer circular solutions for retrofitting purposes and support long‑term environmental goals. The outputs on recycled aggregates will be based on experimental studies examining the use of conductive repair material with (1) rapid curing techniques and (2) load monitoring in crack areas after repair.