ARCA²D: Detection, analysis and monitoring of deterioration in masonry structures

Each major port authority in France is responsible for maintaining its own critical infrastructure.

What possibilities do drones and Artificial Intelligence offer to accelerate the inspection of masonry structures, such as the “simplified compared visit” (put in place by CEREMA)?

Our SaaS solution

To optimise the management of the engineering structure, we construct its 3D digital twin based on aerial drone scans.

In parallel, the acquired images are automatically analysed to detect anomalies (movements, cracks, fractures, separations, etc.) with millimetre precision, geolocate them with centimetre precision and quantify them for ongoing monitoring.

The digital twin becomes the platform for visualising defects, accessible to operators and maintenance and engineering teams, improving information sharing among all stakeholders. In addition to centralising observations and interpretations in this shared 3D model of the structure, digitalisation allows pooling of annotations from various experts in order to enhance our Deep Learning algorithms’ learning bases.

Note that raw data is always available to experts to substantiate their doubt resolution work.

Client benefits

Time savings during acquisition, delivery and monitoring through the digitalisation of the entire process.

Increased personal safety through remote acquisition.

Reduced carbon footprint and contribution to the UN Sustainable Development Goals

In summary

Automatically, based on drone RTK captures, our application makes it possible to:

obtain the 3D digital twin of the structure and explore it virtually,
identify potential defects and geolocate them on the digital twin,
determine the evolution of each defect compared to a previous capture,
provide an immediate summary view of areas of interest,
empower the user to validate and annotate defects highlighted by our AI,
produce data extractions for maintenance operations.

In October 2022, this solution won the Digital Innovation award at the Port du Futur Conference.

With A2D, we developed this decision support software tool for the automatic analysis of deteriorations in maritime infrastructures. The time saving is immense, as teams no longer need to carry out time-consuming inspections by boat, as the system only sends them predefined alerts. Managers can then decide to verify the information and implement a maintenance plan. This significantly reduces travel and, consequently, CO2 emissions, while enhancing employee safety. Digitalisation improves information sharing among all participants, allowing the centralisation of observations and interpretations on a shared digital twin of the structure, as well as experts’ ability to annotate data in order to strengthen Machine Learning algorithm bases.