TY - GEN A1 - Assafo, Maryam A1 - Städter, Jost Philipp A1 - Meisel, Tenia A1 - Langendörfer, Peter T1 - On the Stability and Homogeneous Ensemble of Feature Selection for Predictive Maintenance: A Classification Application for Tool Condition Monitoring in Milling T2 - Sensors N2 - Feature selection (FS) represents an essential step for many machine learning-based predictive maintenance (PdM) applications, including various industrial processes, components, and monitoring tasks. The selected features not only serve as inputs to the learning models but also can influence further decisions and analysis, e.g., sensor selection and understandability of the PdM system. Hence, before deploying the PdM system, it is crucial to examine the reproducibility and robustness of the selected features under variations in the input data. This is particularly critical for real-world datasets with a low sample-to-dimension ratio (SDR). However, to the best of our knowledge, stability of the FS methods under data variations has not been considered yet in the field of PdM. This paper addresses this issue with an application to tool condition monitoring in milling, where classifiers based on support vector machines and random forest were employed. We used a five-fold cross-validation to evaluate three popular filter-based FS methods, namely Fisher score, minimum redundancy maximum relevance (mRMR), and ReliefF, in terms of both stability and macro-F1. Further, for each method, we investigated the impact of the homogeneous FS ensemble on both performance indicators. To gain broad insights, we used four (2:2) milling datasets obtained from our experiments and NASA’s repository, which differ in the operating conditions, sensors, SDR, number of classes, etc. For each dataset, the study was conducted for two individual sensors and their fusion. Among the conclusions: (1) Different FS methods can yield comparable macro-F1 yet considerably different FS stability values. (2) Fisher score (single and/or ensemble) is superior in most of the cases. (3) mRMR’s stability is overall the lowest, the most variable over different settings (e.g., sensor(s), subset cardinality), and the one that benefits the most from the ensemble. KW - classification KW - feature selection KW - homogeneous feature selection ensemble KW - predictive maintenance KW - milling KW - sensor fusion KW - stability of feature selection KW - tool condition monitoring Y1 - 2023 U6 - https://doi.org/10.3390/s23094461 SN - 1424-8220 VL - 23 IS - 9 ER - TY - GEN A1 - Bilous, Vadym A1 - Sarachuk, Kirill T1 - Are SMEs Ready for AI Embedded Mobile Robots? T2 - Human Interaction and Emerging Technologies N2 - The introduction of both mobile robots (MR) and AI-embedded mobile robots (AIMR) into the industry is very slow compared to other types of industrial robots (IR) and automation systems. Many scientific articles and studies are focused on the programming and design of MR. At the same time, integration issues, topical problems and related obstacles are almost entirely absent from the scientific literature. The authors of this paper acknowledge that the complete analysis of this area is a very challenging task. Therefore, for the purposes of this study, we focus on the local problem of analysing the introduction of MR and AIMR in small and medium-sized enterprises (SMEs). The authors offer the analysis of the current challenges and trends in the introduction of mobile robots into SMEs. They also propose solutions to these problems based on their own as well as external experience in the design, programming and implementation of mobile robots. Y1 - 2023 UR - https://openaccess.cms-conferences.org/publications/book/978-1-958651-87-2/article/978-1-958651-87-2_32 U6 - https://doi.org/10.54941/ahfe1004035 VL - 111 / 2023 SP - 285 EP - 296 PB - IHIET 2023 ER - TY - GEN A1 - Bilous, Vadym A1 - Sarachuk, Kirill T1 - Can small and medium enterprises benefit from AR technology? Current challenges and trends T2 - Human Interaction & Emerging Technologies N2 - Albeit augmented reality (AR) technologies first have been discovered in the third quarter of the 20th century, their widespread use began just two decades ago. Existing paper trails show that AR has a wide range of industrial application: it simplifies human-machine communication, improves human-machine interfaces (HMI) for quick information exchange in training (including feedback to study the workflow), correction of errors, machine maintenance, assembly assistance etc. However, broader industrial acceptance of AR, prior to all by small and medium-sized enterprises (SMEs), recently faced considerable problems and the expansion of AR solutions does not match the high potential it has demonstrated. That results in a limited practical use, mainly for demonstration and advertising purposes. This short review is to present the state of the art of the industry, challenges that SMEs face in adopting AR technologies, and several practical examples of a (commercial) adoption of AR. Some prospects for further development of AR and its ongoing integration into industry are briefly discussed in the summary. Y1 - 2023 UR - https://openaccess.cms-conferences.org/publications/book/978-1-958651-87-2/article/978-1-958651-87-2_37 U6 - https://doi.org/10.54941/ahfe1004040 VL - 111 / 2023 SP - 333 EP - 348 PB - IHIET 2023 ER - TY - GEN A1 - Arnouk, Annamria A1 - Behm, Martin A1 - Borck, Christian ED - Yurish, Sergey Y T1 - Architecture for digital twins in manual assembly within the industrial metaverse T2 - Automation, Robotics & Communications for Industry 4.0/5.0 : proceedings of the 6th Winter IFSA Conference on Automation, Robotics & Communications for Industry 4.0/5.0/6.0 (ARCI' 2026) N2 - This paper presents a modular digital twin framework tailored for manual assembly processes within the broader context of the industrial metaverse. Leveraging the NVIDIA Omniverse ecosystem and the Universal Scene Description (USD) format, the architecture unifies design and operational data through sensor-driven inference and real-time extended reality (XR) feedback. The focus is placed on integrating a part of a worker guidance system for detecting assembly actions, defining and controlling USD SubLayers for up-to-date representations of the assembly state, enabling adaptive operator support and establishing closed-loop systems to enable continuous process optimization. The study lays a foundation for scalable deployment across high-variability production environments. KW - Life cycle KW - Digital thread KW - Hybrid-electric aviation KW - Manual assembly KW - Extended reality KW - Process animation Y1 - 2026 SN - 978-84-09-82030-6 U6 - https://doi.org/10.13140/RG.2.2.34764.27526 SN - 2938-4796 SP - 5 EP - 12 PB - IFSA Publishing, S. L. CY - Barcelona, Spain ER -