The optimization of business processes is essential for enhancing efficiency and competitiveness in today’s dynamic industrial environment. This study presents a novel approach that integrates Artificial Intelligence (AI) with Failure Mode and Effects Analysis (FMEA) to automate risk assessment and enable predictive maintenance. Using real-world data from a glass manufacturing company (SK), we implemented machine learning models—Random Forest and XGBoost—to predict Risk Priority Number (RPN) values and classify failure risks. The Random Forest Regressor achieved an Coeficient of Determination (R²) of 0.999, Mean Absolute Error (MAE) of 2.18, and Root Mean Square Error (RMSE) of 2.65, indicating high accuracy in risk prediction. The Random Forest Classifier reached a classification accuracy of 82.9%, with strong recall (92.4%) for high-risk cases, making it practical for industrial deployment. The proposed AI-enhanced FMEA system supports dynamic risk management, reduces downtime, and improves maintenance planning. Our findings demonstrate that AI can significantly strengthen traditional FMEA practices and serve as a foundation for intelligent, data-driven decision-making in industrial environments.

This article introduces an innovative inspection system designed to check the condition of the welding electrode tip. The system combines 2D laser scanning with advanced analytical algorithms, enabling fast, accurate, and automated inspection. Developed PC software controls the synchronized movement of welding electrodes and the scanner, allowing the creation of 3D point clouds representing the surface of the electrodes. These data are analysed to identify deformations, contamination, and surface imperfections. The inspection results are communicated with the welding control system, contributing to improved weld quality and efficiency in industrial production processes. This innovation has the potential to change how the condition of welding electrodes is monitored in industrial sectors.

This paper aims to describe a method of non-destructive testing in highly variable production using a robotic ultrasound diagnostic system. Highly variable production involves producing products of different shapes and dimensions, which requires flexible positioning and the adaptation of diagnostic technology. Typically, highly variable products are made in small batches. Implementing automation tools, including robotic systems, can meet this requirement, even in diagnostic processes that demonstrate the required quality level. This paper deals with the design, simulation, optimization, and verification of an innovative robotic ultrasound diagnostic system for non-destructive testing in highly variable production environments. The robotic positioning of the diagnosed object and the ultrasound probe achieves the automatic adaptation of the system. Robots can automatically position the ultrasound probe relative to the diagnosed object using laser distance measurements in the water environment of a diagnostic vessel. Computer processing of the data measured by the ultrasound probe enables the evaluation of data and the documentation of quality criterion fulfillment in digital form. Setting the technology parameters, monitoring the technology status, and displaying the quality control results are enabled by a human–machine interface system also described in this paper.

This paper presents an investigation of the viability of Fiber Bragg Grating (FBG) accelerometers and load sensors for two diagnostic tasks in robotics: vibration analysis of industrial manipulators and slip detection in robotic grippers. The study synthesizes experimental results from two independent experiments, focusing on the use of optical sensors in place of traditional electromechanical devices such as MEMS accelerometers and resistive strain gauges. The experiments were performed on an ABB industrial robot with a parallel robotic gripper equipped with embedded optical fibers. Under the conditions tested, the optical chains provided good sensitivity and detected dynamic events such as slip onset and contact or movement-induced vibrations. Compared with the conventional chains recorded alongside them, the fiber-optic accelerometers delivered cleaner vibration spectra (SNR 36.7 dB against 13.7 dB), while the FBG load sensors measured gripping forces with a lower noise floor and less mains-band interference, although the force lost before a slip alarm favored the FBG chain only in the experiment carried out with brass. Obtained with a single specimen of each sensor type and a limited set of grasped objects, these findings indicate that optical sensing is a viable alternative for robotic diagnostics, but it comes at the cost of higher price and complexity.