In-situ monitoring of geometrical defects in laser powder bed fusion: probability of detection across defect sizes and sensor architectures

In-situ sensing and monitoring have been extensively researched for detecting anomalies and defects of industrial relevance, while the part is being built. Such capability has been pointed out to have a key role in the design and development of more efficient product qualification and process verification practices across various application sectors. Several methods have been proposed in the literature. Among them, those focused on the in-situ detection of dimensional and geometrical distortions in laser powder bed fusion are currently characterized by a quite high technology readiness level. This makes them particularly attractive for adoption in real production environments. However, these methods have often been validated under distinct experimental conditions, complicating the assessment of their transferability, scalability, and generality. This paper presents the first systematic study on the impact of defect size on the probability of detection for such family of in-situ monitoring techniques. By utilizing a range of in-situ sensing architectures—encompassing different sensors, illumination settings, and L-PBF machines—the study aims to provide a more comprehensive understanding of defect detectability across various industrial setups. The findings contribute to identifying the critical parameters and architectural choices that influence the in-situ detection performance, facilitating the selection of most appropriate solutions for industrial implementation.