Written by Julian von der Goltz

At Mainblades, our mission is to revolutionize aircraft maintenance by significantly reducing inspection time while elevating inspection quality. In this article, we will explore the profound advantages that Artificial Intelligence (AI) brings to aircraft inspections.

Before we go into the “AI” part, let us first examine the physical or “real-world” circumstances of a drone inspection:

1. The drone flies at a fixed distance from the aircraft surface. The distance is set as a parameter before the flight when choosing the inspection type. A typical value would be around 2.5 meters. A full-body inspection then typically takes between 45-60min for narrowbody and 2-3h for a widebody aircraft.
2. We utilize high-quality cameras like the Sony A7R. This full frame camera sensor has a size of 35.7x23.8 mm that maps to 9504x6336 pixels.
3. In addition to the camera, we use a lens with a high focal length, that allows us to get incredibly detailed images of the aircraft surface. A typical focal length is 50 mm.

Let us simplify the technical complexities: our drones fly at a predefined distance from the aircraft, capturing detailed images with ground sampling distances as small as 0.2mm/px. This means we can detect objects as small as 1-2mm on aircraft as large as the Boeing 777–300.

But what does this mean in practice? Compared to manual inspections, AI-powered drone inspections offer unparalleled efficiency and accuracy. Our AI models meticulously analyse every inch of the aircraft surface, identifying patterns and anomalies indicative of damage or defects. This rigorous analysis ensures that no detail goes unnoticed, mitigating the risk of human error or oversight.

In the second part of our detection pipeline, our system makes intelligent suggestions for the operator to focus on, based on the inspection type. The inspection types are customized for each use case. For example, a general visual inspection or an inspection for paint damages or for lightning strikes.  

‍

‍

We use a combination of specialized expert AIs for each use case, the area that is being inspected, and the severity of the findings to automatically filter out insignificant findings and highlight only the relevant areas of concern for the human operator to review.

For a general visual inspection, our AI can tag cracks, dents, corrosion, missing rivets/fasteners, and other structural defects with high accuracy. For paint damage inspections, it can identify scratches, peeling, discoloration, and paint chipping areas. In the case of lightning strike inspections, it can pinpoint burn marks, surface dents and composite delamination caused by lightning strikes.

The AI’s ability to automatically recognize and classify these defects saves countless hours compared to manual visual inspection methods. Instead of spending many hours on a manual inspection, the operator only needs to verify the AI suggestions in one of our apps. Our user interfaces are designed with this human-AI-interaction in mind and offer a fast review experience.

The consistent, objective nature of AI detection means no defect goes unnoticed due to human fatigue or error. Our AI works tirelessly, applying the same rigorous analysis to every inch of the aircraft surface.

In conclusion, AI-powered drone inspections represent a transformative leap forward in aircraft maintenance. By harnessing the precision of AI technology and the agility of drone platforms, we can conduct comprehensive inspections with unparalleled speed and accuracy. As the aviation industry embraces automation, these innovative solutions will become indispensable tools for ensuring safety, efficiency, and regulatory compliance in aircraft maintenance.

At Mainblades, we are proud to lead the charge in this technological revolution, shaping the future of aircraft maintenance one inspection at a time.

‍

‍

Mainblades

www.mainblades.com
marketing@mainblades.com

‍