The high initial costs in hi-tec inspection technology and AI forecasting will pay off in time savings and reduced maintenance costs, the PAM APAC conference was told.
Anthony Mannion, principal research fellow at Aviation Services Research Centre at Hog Kong Polytechnic University, presented some of pioneering work it is doing.
This includes using the latest hyperspectral drone inspection technology to give engineers a much more detailed view of damage and corrosion of aircraft parts.
This high spectral imaging, which sees in 218 colours compared to the human eye which just sees in three, is combined with electrochemical signals to produce early warning of degradation.
Combined with geo-location data the system creates a virtual simulation of the airframe that identifies precisely where all the areas of degradation are and saves all the meta-data related that inspection.
Artificial intelligence models can then be used to predict the likely progress of that damage and inform airlines when they need to intervene to fix it.
A Large Language Model can be used by engineers to query the inspection data and predictive future requirements for repair, replacement parts, and labour.
Another area the university is working on is pre-delivery status reports on engines. This ‘You Only Look Once’ inspection technology maps the engine and highlights in a report for the operator any areas of concern.
Radar inspection and ongoing assessment of aircraft radomes is also an important practical use of this technology due to the likelihood of damage of this vulnerable part of the airframe.
Mannion said generating large amounts of data and using AI to scan it will help airlines to “move from reactive maintenance to predictive maintenance”.
“Initially there are high investment costs but generally it will come back because of the advantage of time saving and maintenance costs,” he said.
