Smart Automated Aircraft Visual Inspection System (SAAVIS)

The Smart Automated Aircraft Visual Inspection System (SAAVIS) is a novel defects detection system that utilises overhead high- resolution cameras (with AI-based computer vision capability) and autonomous ground robots (with 3D localisation capabilities) to detect various defects on the airframe.



Example of the SAAVIS overhead cameras and mobile robotics system

Airframe-centric navigation robotic system with telescopic arms for capturing images at various heights around the airframe and localising them with respect to the aircraft.

Its enhanced autonomous mobile robot uses 3D dome-lidar technology to localise and autonomously navigate itself with respect to the aircraft as an anchor, instead of relying on the surrounding environment. Therefore, the autonomous robot can visually inspect the airframe at any location (e.g. hangar or tarmac), without the need to have a pre-scanned map of the area.  Additionally, the robots are equipped with collision avoidance capability to prevent accidents with their surroundings and people.

This can improve the accuracy and speed of defect detection during pre-flight checks, and airplane maintenance processes on airframe surfaces like wings and undercarriage. 

Hybrid-AI computer vision system that recognises a wide range of defects using less data. 

There are over 50 different types of airframe defects, but commonly used deep learning (DL) methods require extensive image datasets for training each defect class, and high-resolution images of these defects are rare. 

Therefore, we adopted a Hybrid-AI method in our defects recognition system. Combining traditional machine-learning with deep-learning methods, SAAVIS can identify more than 30 commonly known types of defects, such as lighting strikes, dents, scratches, and loose screws, with minimal training data. 

Usual aircraft inspections are manual and require work at height. With this technology, it reduces the defect detection time and the risk of defects being overlooked during manual human inspections. 

Contact us to know more about industry collaboration opportunities with us.