Ships must be inspected for the classification and business operation. However, it is difficult to find a balance between regular inspections and inspection costs, so as to avoid corrosion, serious damage or even catastrophic errors on the ships.
Until now, drone inspection attempts required human pilots and specialists, and, opposite to the classification inspection, they have only been based on visual input. Drones now aspire to carry out this task, to improve precision.
The 'Inspectrone project' aims to develop an autonomous system that will use many different sensors, providing consistent and regular inspection data. It is expected that the technology will be used in various areas that require inspection of closed, dangerous and other hard-to-access areas.
With the help of Artificial Intelligence, the Technical University of Denmark will find errors in images, and through touch, technology can let drones test their surroundings.
In addition, the project wants to establish a user-friendly system that does not require expert users, and which can provide objective and accurate inspections on its own.
"Our system contains human knowledge, but is, at the same time, more precise and objective. It will revolutionise ship inspections. The system is not costly and does not require an expert’s input, as simply pressing a button on a screen is required. The system can carry out regular inspections and provide vital information on the health status of ships."
highlighted Evangelos Boukas, Assistant Professor at the Technical University of Denmark.