Laser beam powder bed fusion (PBF-LB) is a widely adopted metal additive manufacturing (AM) technique in which a high-power laser selectively melts and fuses metal powders layer by layer. Despite significant technological advances, PBF-LB still suffers from limited process robustness and repeatability, often resulting in surface imperfections that degrade part quality. While macroscale defects such as recoater-induced damage or macrocracks are relatively easy to identify, microscale imperfections, including balling, spattering and surface porosity, remain difficult to detect in real time, yet can critically affect internal part integrity. Conventional inspection methods struggle to capture these fine-scale defects, highlighting the need for advanced, non-destructive approaches for in-situ defect detection. This work evaluates the performance of pretrained, unsupervised machine learning (ML) models for detecting and characterising microscale surface imperfections on PBF-LB surfaces using light scattering. The pretrained ML algorithms include an autoencoder and an anomaly detection model (Isolation Forest) trained on physics-based simulations of scattering patterns from an AM surface. The models are applied directly to experimental scattering patterns from AM surfaces with varying topographies and materials, allowing assessment of model generalisation. This approach shows how variations in surface topography influence scattering behaviour and defect detectability. By analysing parts fabricated under different PBF-LB processing parameters, the study evaluates the applicability of the trained ML model for assessing the quality of additively manufactured surfaces with varying topography characteristics.

Fourier transform-based frequency sweeping interferometry (FT-FSI) is an interferometric technique that enables absolute distance measurement by detecting the beat frequencies from the interference of reflected signals. This method allows robust, simultaneous distance measurements to multiple targets and is largely immune to variations in the reflected optical signal intensity. As a result, FT-FSI maintains accuracy even when measuring reflectors with low reflectance. FT-FSI has recently been integrated into the full remote alignment system (FRAS) developed for the High-Luminosity Large Hadron Collider (HL-LHC) project at CERN. Designed to operate in harsh environments with electromagnetic interference, ionizing radiation and cryogenic temperatures, FRAS employs FT-FSI for the precise monitoring of the alignment of accelerator components. The system includes specialized interferometers and a range of sensors, including inclinometers, distance sensors, and leveling sensors. This paper presents a comprehensive review of the challenges associated with remote measurement and monitoring systems in harsh environments such as those of particle accelerators. It details the development and validation of the FT-FSI-based measurement system, emphasizing its critical role in enabling micrometric alignment accuracy. The developments and results presented in this work can be readily translated to other demanding metrology applications in harsh environments.

152Dimensional metrology is one of the cornerstones of mechanical engineering and is key to manufacturing. Dimensions have to be confirmed by measurement, and such measurements allow tolerances and uncertainties to be calculated. Following on from the metrology basics covered in Chapter 2, this chapter will present the multitude of instruments and techniques that are available for dimensional metrology, including the measurement of length, three-dimensional coordinates, and surface form and texture. Single and multi-dimensional measuring instruments will be discussed along with their capabilities and limitations. Reversal methods will also be covered.

Microscale surface imperfections in laser beam powder bed fusion (PBF-LB) additively manufactured parts, such as balling, spattering, and surface pores, can substantially reduce component quality but are difficult to detect with current real-time measurement and monitoring methods. This paper introduces a novel, rapid, and costeffective method for detecting microscale surface imperfections in PBF-LB, utilising light scattering combined with machine learning (ML) algorithms. In the proposed method, a laser beam illuminates the measured surface, and the scattered light is captured and analysed to detect surface imperfections. The scattering patterns, which are associated with the illuminated surface and the configuration of the setup, are used to train unsupervised ML algorithms, including autoencoders and anomaly detection models, to classify surfaces as either uniform, without any imperfections or non-uniform, with imperfections. The ML models were trained on simulated scattering patterns of synthetic surfaces generated by a generative adversarial network (GAN) and validated on experimental datasets. The use of unsupervised models eliminates the need for data labelling, whilst the use of simulated and synthetically generated data reduces the time required for actual experiments and data collection. Experimental validation demonstrates that the most effective trained ML model achieved a classification accuracy of over 97 %, highlighting the potential of this technique for detecting microscale surface imperfections. This paper demonstrates the capability of our method to detect such imperfections on PBF-LB surfaces as an exsitu process. Nonetheless, with further development, this approach has the potential to be adapted as on-machine and real-time defect detection method, by integrating the illumination source into a commercial PBF-LB machine and capturing scattered light information for real-time quality monitoring during the manufacturing process.

Rigorous scattering models are based on Maxwell's equations and can provide high-accuracy solutions to model electromagnetic wave scattering from objects. Being able to calculate the scattered field from any surface geometry and considering the effect of the polarisation of the incident light, make rigorous models the most promising tools for complex light-matter interaction problems. The total intensity of the electric near-field scattering from a silicon cylinder illuminated by the transverse electric and transverse magnetic polarisation of the incident light is obtained using various rigorous models including, the local field Fourier modal method, boundary element method and finite element method. The intensity of the total electric near-field obtained by these rigorous models is compared using the Mie solution as a reference for both polarisation modes of the incident light. Additionally, the intensity of the total electric near-field scattered from a silicon sinusoid profile using the same rigorous models is analysed. The results are discussed in detail, and for the cylinder, the deviations in the intensity of the total electric field from the exact Mie solution are investigated.

Uncertainty evaluation in the measurement of surfaces with complex topography using optical techniques remains a challenge due to the complex interaction between light and the surfaces. The ISO 25178 part 600 specification standard simplifies uncertainty evaluation by introducing a set of agreed metrological characteristics that can be propagated through a mathematical model. To complement this, we developed a virtual coherence scanning interferometer to model error sources and provide task-specific uncertainty evaluation. This paper presents a framework for evaluating measurement uncertainty of areal surface texture parameters, using both the metrological characteristics approach and the virtual instrument method. We demonstrate this framework by assessing the uncertainty of the Sq parameter, which represents the root-mean-square of surface heights, for sinusoidal and quasi-random surfaces. By comparing the combined standard uncertainty from both approaches, we quantify the contribution of topography fidelity, a key but difficult-to-evaluate characteristic. The proposed method offers a comprehensive understanding of uncertainty in optical surface measurement, leading to improved tolerancing in manufacturing.

Coherence scanning interferometry (CSI), based on the principle of interference, can achieve sub-nanometer precision for height measurements. On the other hand, focus variation microscopy (FVM), combining the small depth of field of the objective, is a widely used surface topography measurement method suited to surface topography that is mostly optically rough. In this paper, we propose a method to simultaneously obtain the interferometric fringe data and focus variation FVM image stack, from a single vertical scanning process, using a CSI instrument without any hardware modifications. Using a 3D Fourier transform, the FVM signal, looks takes the form of a “bowtie” and the CSI signal resembles two “umbrellas” that are separated in 3D K-space. The signal is recovered using a 3D inverse Fourier transform and the surface topography can be determined by fusing the CSI and FVM signals. Since both signals come from the same instrument and scanning process, there is no need for coordinate registration and data interpolation during the data fusion process. Our method combines the features of CSI and FVM measurement, thereby improving the robustness and data coverage of the measurement. An all-in-focus surface topography map can also be generated using this method. This focusing feature has the potential to significantly improve the defect detection and quality control ability of CSI instruments.

In optical metrology, the growing demand for accurate measurement technologies is driven by the increasing applications of three-dimensional (3D) microscopy and imaging. The advancement of these technologies relies on the modelling of the measurement process, where an initial step involves characterising the interaction between electromagnetic fields and surfaces, to determine the scattered electromagnetic field, and the subsequent propagation of the scattered light through the instrument. Virtual instruments (VIs) can play a critical role by replicating optical instruments through surface scattering models, 3D imaging theory and error-generation models. The development of VIs contributes to a better understanding of instrument characteristics, optimising configurations and evaluating uncertainties. VIs can be customised to simulate various optical setups, providing researchers with flexibility to explore optimal measurement settings. Furthermore, the integration of sophisticated computational tools and machine learning to VIs can enable real-time, in-depth and optimal analysis of optical instrument modelling. Coherence scanning interferometry (CSI) is a widely used optical technique for high-accuracy surface topography measurement. A virtual CSI (VCSI) models the CSI measurement process using physics-based models. Approximate scattering models, utilising basic scalar diffraction and linear imaging theories, facilitate CSI modelling that can offer insights into the fundamental sources of measurement error for smooth surfaces. However, with the increasing use of complex freeform manufactured structures in engineering, aerospace and biology, scrutinising complex surface features becomes more critical. As a result, rigorous scattering models, based on solutions of Maxwell's equations, can be used as tools for addressing more complex light-matter interactions. Capable of calculating light scattered fields from any surface geometry while taking into account different light phenomena, such as polarisation, rigorous models provide high-accuracy solutions. In this study, we compare the simulated fringe pattern and reconstructed profiles obtained by virtual coherence scanning, employing both approximate and rigorous scattering models for sinusoidal and vee-groove samples.

Coherence scanning interferometry (CSI) is a widely used optical method for surface topography measurement of industrial and biomedical surfaces. The operation of CSI can be modeled using approximate physics-based approaches with minimal computational effort. A critical aspect of CSI modeling is defining the transfer function for the imaging properties of the instrument to predict the interference fringes from which topography information is extracted. Approximate methods, for example, elementary Fourier optics, universal Fourier optics, and foil models, use scalar diffraction theory and the imaging properties of the optical system to model CSI surface topography measurement. In this work, the simulated topographies of different surfaces, including various sinusoids, two posts, and a step height, calculated using the three example methods are compared. The presented results illustrate the agreement between the three example models. (c) The Authors. Published by SPIE under a Creative Commons Attribution 4.0 International License.Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI

New developments in vision algorithms prioritise identification and perception over accurate coordinate measurement due to the complex problem of resolving object form and pose from images. Consequently, many vision algorithms for coordinate measurements rely on known targets of primitive forms that are typically planar targets with coded patterns placed in the field of view of vision systems. Although planar targets are commonly used, they have some drawbacks, including calibration difficulties, limited viewing angles, and increased localisation uncertainties. While traditional tactile coordinate measurement systems (CMSs) adopt spherical targets as the de facto artefacts for calibration and 3D registration, the use of spheres in vision systems is limited to occasional performance verification tasks. Despite being simple to calibrate and not having orientationdependant limitations, sphere targets are infrequently used for vision-based in-situ coordinate metrology due to the lack of efficient multi-view vision algorithms for accurate sphere measurements. Here, we propose an edge-based vision measurement system that uses a multi-sphere artefact and new measurement models to extract sphere information and derive 3D coordinate measurements. Using a spatially encoded sphere identities embedded in the artefact, a sphere matching algorithm is developed to support pose determination and tracking. The proposed algorithms are evaluated for robustness, measurement quality and computational speed to assess their performance. At the range of 500 mm to 750 mm, sphere size errors of less than 25 mu m and sphere-to-sphere length errors of less than 100 mu m are achievable. In addition, the proposed algorithms are shown to improve robustness by up to a factor of four and boost computational speed.

The implementation of in-situ sensing solutions to monitor additive manufacturing processes has seen a significant surge in recent years, notably in processes where localised heating is used. These technologies, however, have not always yielded accurate information about final part quality, due to differences between the irregularities observed in-process and the anomalies present in the finished part. One way of investigating such differences is to establish correlations between in-process layer properties and the final condition of the part. In this work, we put forward a solution based on a bespoke fringe projection system designed to monitor layers within the build chamber of a PBF-LB machine. Through the computation of quantitative indicators on fringe projection data and the use of statistical control charts for their monitoring and analysis, we are able to predict local reductions in part density, detrimental to the quality of the final build, which are typically only visible after part manufacture. Principally, this article describes the developed fringe projection system utilised for data collection and the custom indicators used to examine layer topographical characteristics and which are correlated with local final densities. We design an experimental campaign to produce parts with different local densities and show how the proposed indicators, combined with statistical control charts, can predict in-process drops in density. The monitoring performance is validated via X-ray computed tomography (XCT) measurements performed on the as-built samples.

Abstract Despite ongoing improvements and optimisation efforts, the powder bed fusion (PBF) process continues to face challenges related to repeatability, robustness, and stability. These challenges can lead to the formation of microscale surface impurities on each layer, such as balling, spatter and surface pores, which can adversely affect the overall quality of the final part. The layer-by-layer fabrication approach in PBF offers an opportunity to assess fabrication quality in real-time by detecting these impurities at each layer during the manufacturing process through in-situ sensing methods. With advancements in sensing and computing technologies, there has been a significant increase in studies focused on developing in-situ methods for the real-time detection of surface impurities and feedback mechanisms. However, it is necessary to understand the effectiveness and capability of these in-situ methods in detecting microscale surface impurities, as well as to evaluate their potential advantages, drawbacks, and the existing gaps in the literature. This study first summarises the common microscale surface impurities and their potential impacts on part quality, including mechanical properties and surface finish. It then reviews the existing in-situ methods capable of detecting these microscale impurities, providing insights into the strengths and limitations of current techniques, and identifying gaps in the literature while suggesting directions for future research.

Abstract This literature review investigates the integration of machine learning (ML) into optical metrology, unveiling enhancements in both efficiency and effectiveness of measurement processes. With a focus on phase demodulation, unwrapping, and phase-to-height conversion, the review highlights how ML algorithms have transformed traditional optical metrology techniques, offering improved speed, accuracy, and data processing capabilities. Efficiency improvements are underscored by advancements in data generation, intelligent sampling, and processing strategies, where ML algorithms have accelerated the metrological evaluations. Effectiveness is enhanced in measurement precision, with ML providing robust solutions to complex pattern recognition and noise reduction challenges. Additionally, the role of parallel computing using graphics processing units and field programmable gate arrays is emphasised, showcasing their importance in supporting the computationally intensive ML algorithms for real-time processing. This review culminates in identifying future research directions, emphasising the potential of advanced ML models and broader applications within optical metrology. Through this investigation, the review articulates a future where optical metrology, empowered by ML, achieves improved levels of operational efficiency and effectiveness.

As the need for the manufacturing of complex surface topographies increases, traceable measurement with known uncertainties can allow a manufacturing process to remain stable. Material measures are the link in the chain that connects the surface topography measurement instrument’s output to the definition of the metre. In this review, the use of material measures is examined for the purposes of instrument calibration and performance verification based on the metrological characteristics framework, as introduced in ISO 25178 part 600. The material measures associated with each metrological characteristic are investigated in terms of fabrication, geometry and functionality. Material measures for metrological characteristics are discussed in a sequential approach, focusing on material measures that have been developed for specific measurement technologies and optical surface topography measurement instruments. There remains a gap in the metrological characteristic framework for the characteristic, topography fidelity, and the review highlights current methods using reference metrology and alternative approaches using virtual instruments to quantify the effects of topography fidelity. The influence of primary instruments is also reviewed in the context of uncertainty propagation. In the conclusion, the current challenges are identified with regards to the scarcity of available material measures in the lower nanometre range, and the limitations in terms of cost, complexity, manufacturing time and industrial applicability.

The majority of industrial production processes can be divided into a series of object manipulation and handling tasks that can be adapted for robots. Through significant advances in compliant grasping, sensing and actuation technologies, robots are now capable of carrying out human-like flexible and dexterous object manipulation tasks. During operation, robots are required to position objects within tolerances specified for every operation in an industrial process. The ability of a robot to meet these tolerances is the critical deciding factor that determines where the robot can be integrated and how proficient the robot can carry out high-precision tasks. Therefore, improving the positioning accuracy of robots can lead to new avenues for their integration into production industries. Given that tolerances in manufacturing processes are in the order of tens of micrometres or less, robots should guarantee high positioning accuracy when manipulating objects. The direct method of ensuring high accuracy is by introducing an additional measurement system(s) that can improve the inherent joint-angle-based robot position determination. In this paper, we present a High-Accuracy Robotic Pose Measurement (HARPM) system based on coordinate measurements from a multi-camera vision system. We also discuss the integration of measurements obtained by absolute distance interferometry and how the interferometric measurements can complement the vision system measurements. The performance of the HARPM system is evaluated using a laser interferometer to investigate robotic positions along a trajectory. The performance results show that the HARPM system can improve the positioning accuracy of robots from hundreds to a few tens of micrometres.

Static friction modeling is a critical task to have the accurate robot model. In this article, a neural network separation approach to include nonlinear static friction in models of industrial robots is proposed. For this purpose, the terms corresponding to static friction within the overall robot mathematical model are separable terms treated independently from the rest of the model. The separation modeling process is accomplished by first determining the mathematical model for the system by excluding the friction terms and estimating its parameter values. This part of the model corresponds to gravitational terms only. Because persistency of excitation is required to maintain high accuracy and avoid singularity in the estimations, data with large variations across multiple joint angles are gathered for estimation purposes and a weighted least squares approach is used. This estimation results in a highly accurate static mathematical model for industrial robots. Results from the weighted least squares estimation are compared with the original least squares estimation, ridge regression, a least absolute shrinkage and selection operator, and an elastic net to show superior performance. After modeling the gravitational terms of the model, a multilayer perceptron neural network is used to identify static friction forces in the model from the experimental data. This is required in the case of a robot with multiple degrees of freedom because the friction of each joint is a function of several other joint angles acting upon it; making the solution complex and difficult to be obtained through other friction modeling methods. The experimental results obtained from a Universal Robots-UR5 demonstrate the high accuracy of the proposed modeling methodology under static conditions, and future work will consider the implementation of dynamic terms to integrate friction forces during movement.

Parametric splines are popular tools for precision optical metrology of complex freeform surfaces. However, as a promising topologically unconstrained solution, existing T-spline fitting techniques, such as improved global fitting, local fitting, and split-connect algorithms, still suffer the problems of low computational efficiency, especially in the case of large data scales and high accuracy requirements. This paper proposes a speed-improved algorithm for fast, large-scale freeform point cloud fitting by stitching locally fitted T-splines through three steps of localized operations. Experiments show that the proposed algorithm produces a three-to-eightfold efficiency improvement from the global and local fitting algorithms, and a two-to-fourfold improvement from the latest split-connect algorithm, in high-accuracy and large-scale fitting scenarios. A classical Lena image study showed that the algorithm is at least twice as fast as the split-connect algorithm using fewer than 80% control points of the latter.

Close-range photogrammetry can be used to reconstruct dense point clouds of an object with very high surface coverage, making it useful for manufacturing metrology tasks such as part inspection and validation. However, compared to competing techniques, data processing times can be slow. In this paper we present a method to autonomously remove the background from the images within a photogrammetric dataset. We show that using masked images directly in the reconstruction results in much lower data processing times, with lower memory utilisation. Furthermore, we show that the point density on the object surface is increased while the number of superfluous background points is reduced. Finally, a set of reconstruction results are compared to a set of tactile coordinate measurements. Reconstructions with the background removed are shown to have a standard deviation in the point to mesh distance of up to 30 µm lower than if the background is not removed. This improvement in standard deviation is likely due to the static background, relative to the object on the rotation stage, causing triangulation errors when points are detected and matched on this background data. The proposed approach is shown to be robust over several example artefacts and can, therefore, be implemented to improve the measurement efficiency and measurement results of photogrammetry coordinate measurement systems.

Manufacturing has recently experienced increased adoption of optimised and fast solutions for checking product quality during fabrication, allowing for manufacturing times and costs to be significantly reduced. Due to the integration of machine learning algorithms, advanced sensors and faster processing systems, smart instruments can autonomously plan measurement pipelines, perform decisional tasks and trigger correctional actions as required. In this paper, we summarise the state of the art in smart optical metrology, covering the latest advances in integrated intelligent solutions in optical coordinate and surface metrology, respectively for the measurement of part geometry and surface texture. Within this field, we include the use of a priori knowledge and implementation of machine learning algorithms for measurement planning optimisation. We also cover the development of multi-sensor and multi-view instrument configurations to speed up the measurement process, as well as the design of novel feedback tools for measurement quality evaluation.

As metal additive manufacturing has been increasingly accepted as a viable method of industrial manufacture, there has been a significant uptake in manufacturers wishing to verify and test their parts through analysis of part surface. However, various studies have shown that metal additive surfaces tend to exhibit highly complex features and, thus, represent a challenge to those wishing to undertake measurement and characterisation. Over the past decade, good practice in metal additive surface measurement and characterisation have been developed, ultimately resulting in the creation of a new standard guide, ASTM F3624-23, which summarises that good practice. Here, we explain the background and rationale for the creation of this standard and provide an overview of the contents of the standard. An example case study is then presented, showing the worked good practice guidance in a metal additive surface measurement and characterisation task, namely, a comparative measurement of an example surface using two different instruments. Finally, considerations for future versions of the standard are presented, explaining the need to develop further good practice for novel instruments and to focus on feature-based characterisation approaches.