Modern manufacturing systems are influenced by the growing complexity of mechatronics, control systems, IIoT, and communication technologies integrated into cyber-physical systems. These systems demand flexibility, modularity, and rapid project execution, making digital tools critical for their design. Virtual commissioning, based on digital twins, enables the testing and validation of control systems and designs in virtual environments, reducing risks and accelerating time-to-market. This research explores the development of digital twin models to bridge the gap between simulation and real-world validation. The models identify design flaws, validate the PLC control code, and ensure interoperability across software platforms. A case study involving a modular Festo manufacturing system modelled in Tecnomatix Process Simulate demonstrates the ability of digital twins to detect inefficiencies, such as collision risks, and to validate automation systems virtually. This study highlights the advantages of virtual commissioning for optimizing manufacturing systems. Communication testing showed compatibility across platforms but revealed limitations with certain data types due to software constraints. This research provides practical insights into creating robust digital twin models, improving the flexibility, efficiency, and quality of manufacturing system design. It also offers recommendations to address current challenges in interoperability and system performance.

The efficient and sustainable design of robotic manufacturing cells is a critical aspect of modern industrial processes, for which energy and time efficiency play significant roles in achieving sustainability goals. In industrial practice, robotic cell design often involves methods such as predefined layout templates, empirical rules for positioning, and simulation-based validation. While these approaches provide a practical starting point, they may not fully account for the complex interdependencies between robot configuration, energy consumption, and operational efficiency. Consequently, opportunities for optimizing resource usage are frequently overlooked. This paper presents a novel methodology for optimizing the deployment of industrial robots and their peripherals, focusing on minimizing energy and time costs to enhance the sustainability of industrial processes. The proposed approach, grounded in experimental measurements and simulations, was validated through an experimental model of a welding robot station. The methodology integrates the analysis of the relationship between the robot base position, trajectory, and energy consumption. The results indicate that adjusting the relative positions of robots and work points can achieve energy savings of approximately six percent. Specifically, optimization reduced energy consumption by 1.6731 Wh per work cycle, translating to an annual savings of 0.8794 MWh for a 60 s clock cycle. These findings highlight the practical applicability of the proposed methodology, demonstrating its potential to significantly improve the energy and time efficiency of robotic workplaces.

A digital twin is an effective tool for the design, testing, and validation of control strategies for electro-pneumatic actuators in industrial automation. This study presents and compares three different digital twin models of a pneumatic cylinder with varying levels of physical fidelity—from basic discrete control, through analogue control without pneumatic dynamics, to a complex model simulating pressure, friction, and airflow. The experiments were conducted using the Siemens NX Mechatronics Concept Designer, integrated with the SIMIT emulation platform and a PLC control system via the standardized OPC UA protocol. The main objective was to evaluate simulation accuracy, model flexibility for testing various control strategies, and the ability of the digital twin to reflect changes in PLC algorithms. The results showed that while simple models are suitable for verifying basic logic, only advanced models can realistically replicate the dynamic behaviour of pneumatic systems, including delay phases and pressure influence. A comparison with the experimental study by Jiménez confirmed a strong correlation between the simulated and actual actuator behaviour. In future work, the developed control algorithm will be connected to a physical cylinder to further validate the models and refine control strategies under real-world conditions.

The aim of this thesis is to develop algorithms for configuring tactile sensors in the calibration process. The research involves extensive measurements, data analysis, and assessment of sensor behavior during measurement to evaluate their performance and reliability. A key contribution of this work is the application of progressive diagnostics theory and methods for obtaining new calibration data, along with analytical methods for evaluating sensor reliability. The work includes mathematical calculations necessary for accurate calibration and logical methods for analyzing sensor states, applied in the development of a specialized diagnostic laboratory. The research led to the development of a methodology for analyzing sensor characteristics, reliability, and mathematical corrections. The implemented autonomous calibration algorithm enables automated diagnostics and optimization of sensor lifetime, improving accuracy and operability. Results demonstrate a reduction in measurement deviations from 18.1% to 1.1%, with the tolerance range narrowed to 2.6%. These findings confirm the effectiveness of the methodology in improving sensor calibration and enhancing system reliability. Automated correction mechanisms allow sensors to adapt dynamically to changing conditions, increasing the overall reliability and cost-effectiveness of technological systems. This work provides a comprehensive analysis of sensor calibration and diagnostic methods, focusing on pressure sensors in industrial systems essential for optimizing industrial processes.

This paper presents a model-based feed rate calibration methodology for robotic fused deposition modelling (FDM), developed to improve dimensional accuracy in thin-walled structures. In single-perimeter geometries, where deviations in material deposition directly influence wall thickness and surface quality, precise feed rate control is essential. The proposed computational approach establishes a volumetric relationship between the robot’s motion speed, layer height, and desired extrusion width, enabling accurate determination of the material feed rate from measurable process parameters rather than empirical tuning. The system was implemented within a cyber-physical system (CPS) architecture integrating independent control of the robot and extrusion unit, allowing synchronized motion and material deposition. Experimental validation on cuboid and cylindrical specimens confirmed the model’s effectiveness, achieving minimum wall thickness deviations of 0.027 mm for the cuboid and 0.065 mm for the cylindrical sample. The methodology provides a practical framework for feed rate calibration in robotic FDM systems, supporting consistent extrusion quality, reduced material variation, and faster process setup. Although tested on a specific extrusion system, the approach is scalable to large-format robotic additive manufacturing (RAM) and adaptable to various extrusion materials and configurations.

The use of the sensors shortens the service life, wears out and reduces their accuracy due to operation. For sensors with a susceptibility to inaccuracy, it is possible to create a sensor-device-software diagnostic set. Such a scheme of configuration should be able to provide autonomic diagno...