About xCTing

About xCTing

First-time-right and zero-defect manufacturing of customized lot-size-one products are essential elements of the Industry 4.0 paradigm shift to reinforce Europe’s global leadership in manufacturing. X-ray Computed Tomography (CT) metrology has a key role to play in this transition, since it is the only known technology that can certify non-destructively the quality of internal complex structures, such as those produced by additive manufacturing or found in assemblies.

However, CT largely remains an off-line technology, due to the unsolved trade-off between scan speed and scan quality, and especially the need for extensive expert user input. xCTing will therefore focus on significantly increasing autonomy, robustness and speed in CT metrology in order to support its transition towards a fully in-line quality assurance technology as required in Industry 4.0 environments. Meeting these challenges requires the integration of a broad range of interdisciplinary expertise, including physics, manufacturing, dimensional metrology, machine learning, as well as efficient and reliable big data analytics and visualization.

In order to achieve the envisaged innovation breakthrough in the European industry, Europe is in dire need of young innovators who can combine this variety of competences with entrepreneurial skills. This MSCA ITN project "xCTing" is a pan-European industrial- academic initiative committed to the provision of a unique and encompassing training environment required to foster a new generation of innovation-minded research engineers, that will act as catalysts in the further transformation of Europe’s manufacturing industry towards global technological leadership.

Project objectives

Through the intensive collaborations among the universities, research institutes and industrial stakeholders, the xCTing project consortium expects to bring about R&D breakthroughs in:

  • techniques for autonomous determination of optimal CT scan strategies,
  • deep learning-based reconstruction strategies that correct beam hardening, scatter and other scanning artefacts without expert user input,
  • smart and autonomous segmentation and defect detection algorithms for lot-size-one products by exploiting prior knowledge, such as CAD data in combination with simulations (digital twin) and machine learning techniques applied on historical CT data,
  • a fast Virtual Metrology CT based approach for determining task-specific measurement uncertainty that can even be employed in lot-size-one situations,
  • novel methods to realize autonomous in-scanning characterization, verification, and condition monitoring of the in-line CT equipment,
  • 3D reconstruction techniques that combine high speed and high quality by improving information out of a limited set of in-line projections with a priori information from a high-quality scan of a representative object or from CAD data,
  • using prior information from CAD models to determine the combination of projection angles that yields optimal reconstructions, considering both 360° rotations or just a limited angular range,
  • combination of big data analytics and visualization from in-line CT scanning with process simulations to develop a CT-based autonomous pipeline for AM build preparation, and 
  • a CT metrology assisted manufacturing & assembly pipeline.