Before a new aircraft takes off for the first time, the cabin undergoes a thorough inspection during handover to the airline. A scratch on the armrest, a dent in the cabin wall – tiny imperfections that can become costly: If damages are detected too late, this can not only delay delivery but also lead to compensation claims. A new research project could soon prevent these issues and make visual inspection faster, more efficient, and even possible remotely.
Damage Detection: High-Tech Instead of Manual Labor
Currently, this inspection is carried out by specialists who often have to travel by plane to manually check the cabin over several hours. Instead of labor-intensive manual work, the DrumView* research project relies on AI-supported image recognition, with which the process could work as follows in the future:
- A mobile system moves through the cabin
- Numerous cameras capture every angle
- Software assembles a complete 3D model of the cabin (NeRF)
- AI-based anomaly detection automatically identifies damages
“Snapshot” of the Cabin
The cornerstone of remote inspection is a new machine-learning-based 3D process: Using Neural Radiance Fields (NeRF), the cabin condition is documented as a detailed snapshot. This can be viewed and assessed from any angle remotely. NeRF requires significantly fewer photos than conventional methods to create a precise model and captures reflective surfaces such as windows or displays, which often cause gaps in traditional 3D scan procedures.
Finding What Doesn’t Fit
Damage detection is also accelerated by smart technologies. One of the biggest challenges in AI-based image recognition is collecting enough data for each type of damage to ensure comprehensive detection. DrumView instead relies on anomaly detection, which identifies deviations from the normal cabin structure. This means the AI can detect errors without needing prior definitions. Deviations are quickly identified without requiring specialized training for each damage type.
Goal: More Sustainable Aircraft Systems
The DrumView project kicked off in January 2025 at the ZAL TechCenter and is part of the GATE 2 Roadmap, aiming to promote sustainable aircraft systems through digital development methods and automation to reduce the ecological footprint across the entire product lifecycle.
Are you interested in AI-supported damage detection or have questions about the DrumView project? Share your thoughts with us and give us a call:
Partners
- ZAL GmbH (Consortium Leader)
- 3Daero GmbH
- Institute for Aircraft Production Technology (IFPT) at TU Hamburg
DrumView is funded by the Europäischen Fonds für regionale Entwicklung (EFRE). #EFREhamburg
