ROBOtic System with Intelligent Vision and Control for Tunnel Structural inSPECTion and Evaluation

ROBO-SPECT, driven by the tunnel inspection industry, adapts and integrates recent research results in intelligent control in robotics, computer vision tailored with semi-supervised and active continuous learning and sensing, in an innovative, integrated, robotic system  that automatically scans the intrados for potential defects on the surface and detects and measures radial deformation in the cross-section, distance between parallel cracks, cracks and open joints that impact tunnel stability, with mm accuracies.

This permits, in one pass, both the inspection and structural assessment of tunnels. Intelligent control and robotics tools are interwoven to set an automatic robotic arm manipulation and an autonomous vehicle navigation so as to minimize humans’ interaction.

This way, the structural condition and safety of a tunnel is assessed automatically, reliably and speedily. The initial dataset on tunnel defects is provided from case studies (e.g., from London Underground) to be used not only for transfer learning but also for the evaluation of the structural models.

The robotic system is evaluated and benchmarked at the research infrastructure of tunnels of VSH, at three road tunnels of the Egnatia Motorway and sections of the railway tunnel of London Post Office.

ICCS is the ROBO-SPECT project coordinator, technical manager and responsible for the quality insurance of the project implementation. Apart from the management and administrative roles of ICCS, the I-SENSE team of ICCS will be responsible (together with the French Partner Ecole des Ponts-Paris Tech) for the cognition systems and computer vision aspects of the robotic system. ICCS will be working on the system specifications and design of the vision system able to detect cracks and material degradation intunnels by visual inspection, having as a target to deploy tools able to quickly identify ‘objects’ with a high accuracy(of 1mm) while meeting demanding time constraints.