Tunnel Inspection Water Segmentation (TWS)

The use of artificial intelligence techniques in the management and maintenance of tunnels is essential to increase the safety of the infrastructure and reduce its maintenance costs.

Tunnel Inspection with Artificial Intelligence today still means a lot of manual work.

RMT developed a solution which uses artificial intelligence for automatic damage detection.

The effect of humidity and hydrology in highway and railway tunnels should be monitored accurately to plan maintenance operations efficiently and effectively. Tunnel Water Segmentation (TWS) is the RMT solution addressing the instance segmentation of wet spots.

Implemented borrowing from state of the art deep learning architectures and practices, the algorithm process high resolution data taken from a dedicated scanning system and outputs the associated humidity segmentation map.

Research on the TWS system paves the way for the automation of the defects assessment process for railway and highway infrastructures.

The very high resolution images acquired with the Tunnel Scan system, developed by ADTS, allow an efficient solution to the problem of tunnel maintenance.