Traceable machine vision systems for digital industrial applications: the DI-Vision project

In the context of digital transformation and the EU Green Deal, machine vision systems (MVSs) are emerging as key technologies for automated inspection, quality control, and precision measurement across industries such as aeronautics, pharmaceuticals, medical devices, electronics, and semiconductors. Despite their growing adoption, MVSs often lack traceability and standardised calibration methods, making their deployment in manufacturing indispensable for ensuring accuracy and reliability. The DI-Vision project tackles these challenges by developing traceable material standards with prismatic and complex geometries, better representing real-world industrial surfaces. The project also advances the use of Digital Twins (DTs) to enable robust uncertainty evaluations and accurate MVS modelling. Additionally, dense-image matching algorithms, along with defect detection classifiers and analysis tools, aiming at enhancing the reliability of industrial inspections are proposed. To validate these developments, DI-Vision will investigate industrial case studies across 13 applications, spanning multiple sectors. The project will also deliver Good Practice Guides (GPGs) and establish a measurement infrastructure through collaboration with European Metrology Networks (EMNs) MathMet, TraceLabMed, AdvManu, as well as standards organisations (ISO/TC213, ASTM) and key industrial stakeholders. By providing traceable solutions, DI-Vision will drive advancements in industrial automation and precision manufacturing, ensuring greater reliability and efficiency across diverse sectors.