Research project: HoleListic
Research area: Inspection, data processing, data modeling, XAI
Funded by: Federal Ministry for Economics Affairs and Climate Action
Collaboration with: GmbH, Boeing
Start of the project: June 2023
End of the project: May 2026
Contact person at the institute: M.Sc. Felix Geiger


HoleListic aims to develop a new type of in-line measurement system for drilling based on white light interferometry (WLI), which, in combination with a holistic data model and AI applications, enables high-quality and resource-efficient drilling processes. The measurement system is being developed on the basis of a Boeing use case, but the aim is to transfer the solutions to similar problems in the aircraft environment.

Robust drilling and riveting processes, as the most frequent operation in aircraft production, play a decisive role in the introduction of new aircraft variants. High process stability and quality are essential here, but quality problems lead to a high overall waste of resources (workpieces, tools). A common deficit of previous approaches to determining the quality of drill holes (e.g. monitoring of process forces & acoustic emissions) is the low information content of the test measures, so that there is no direct correlation between the process result in the form of the geometry of the hole and the drill with the process. Due to the difficult accessibility of the drill holes, there is also no suitable measurement technology for this application.

The high-resolution measurement of the actual work result can contribute to an increase in productivity, but only if the sensor is integrated into the automated drilling process at the same time, further determinable variables are recorded and this information is efficiently collected and processed into a holistic data model so that a direct link can be established between the process and the work result.

The IFPT at TUHH is developing the data processing of 3D WLI measurement data and a subsequent interpretation in connection with the quality of the drilling process. Artificial intelligence methods are provided for this purpose, which place special demands on the algorithms to be used due to their data density (size) and their characteristics (3D point clouds). If the evaluated WLI data is linked with other parameters, not only can quality problems be identified, but countermeasures can also be initiated and a self-optimising adaptive drilling process established. For a holistic approach, collected process, environmental and metadata, the measurement data of the production result and, for example, information on tool wear and geometry are integrated into a common model so that it can be used for linked data processing and analysis. The holistic data model is to be extended by methods of explainable AI (XAI) in order to integrate and export information and knowledge into the system via a human-machine interface.

This creates a holistic approach to process-integrated quality assurance, which enables a deep understanding of the process, which is supplemented by further knowledge modules (e.g. models & simulations) and industrial expertise (e.g. integration of WLI in drill head) through networking with the partners of the "Take-Off" programm.

Contact person at the institute: M.Sc. Felix Geiger