Metrological characterisation of Machine Vision Systems for the analysis of surfaces from advanced manufacturing

Surface metrology in advanced manufacturing highlights challenges, such as need of traceability, evaluation of measurement uncertainty and digitalization for an automatized processing of topographies. To this end, Machine Vision systems (MVs), and their accurate representation through Digital Twins, are crucial for addressing these challenges in many high-value industries, enabling increased productivity, enhanced quality control, and reduced waste. Among 3D MVs, Confocal Microscopes, Coherence Scanning Interferometry and Focus Variation instruments are widely used in semiconductor, biomechanical, automobile, energy, pharmaceutical industries. This work addresses the metrological characterisation of machine vision systems used for surface analysis, focusing on the methodologies and challenges associated with quantifying their performance. Key aspects include instrument characterisation according to two ISO standards. ISO 25178-600 specifies the metrological characteristics of areal instruments used for measuring surface topography. These parameters are noise, flatness, amplification of X, Y and Z axes, linearity of X, Y and Z axes, mapping deviation of X and Y axes, spatial resolution, and topography fidelity. The quantification of these parameters is achieved by different material measures under varying experimental conditions, including different objectives, measuring positions, and sample orientations. ISO 25178-700 describes the generic procedures for the calibration, adjustment, and verification of the metrological characteristics of areal topography measuring instruments. The calibration involves the quantification of metrological haracteristics, the adjustment describes the correction of systematic errors, and the verification test the validity of the two previous points. The possible subsequent comparison is the performance specification of instrument characteristics. The outcomes of this study contribute to the development of standardised metrological practices, enabling the robust and reliable deployment of machine vision systems in industrial surface inspection applications. Moreover, we will compare the results from different instruments coming from a university and an NMI and the development of Digital Twins for MVs.