Artificial Intelligence (AI) driven 3D point scanner for monitoring soil plug hazards during the installation of suction caisson foundations

Soil plug hazards pose a significant risk to the successful installation of suction caisson foundations but are currently inadequately monitored using only a single beam echosounder. To address this issue, a new artifi-cial intelligence (AI) driven three-dimensional (3D) point scanner is proposed for monitoring soil plug haz-ards. The proposed scanner is controlled using a Bayesian Optimisation (BO) algorithm, which automatical-ly adapts its data acquisition path in real-time based on previously acquired measurements. Preliminary la-boratory tests were conducted to assess the effectiveness of the proposed scanner. The results showed that the proposed scanner can accurately estimate 3D surfaces with fewer measurement points than a comparable scanner using the conventional scanning method, typically used in existing 3D point scanners. As the pro-posed scanner can estimate the state of the entire surface in much shorter time than existing sensors, it poten-tially offers a more effective method to monitor soil plug hazards.