Abstract

Abstract

Abstract

Abstract
Simultaneous Localization and Mapping (SLAM) algorithms are a key component in enabling autonomous navigation for robotic systems. This study presents a comprehensive assessment of state-of-the-art SLAM algorithms, focusing exclusively on those with Robot Operating System (ROS) support. The study aims to provide insights into the computational performance of these algorithms by leveraging the benchmark results reported in their respective studies. Each algorithm's performance metrics, as reported in their benchmark studies, are analyzed and compared. This comparative analysis not only highlights the strengths and weaknesses of individual algorithms but also provides a broader understanding of their applicability across diverse robotic platforms and environments. Overall, this study contributes to the advancement of SLAM research by offering a comparative evaluation tailored to ROS-supported algorithms. The findings serve as a valuable resource to make informed decisions regarding the selection and implementation of SLAM solutions in real-world applications.

Abstract
The assessment of deposited material in tunnel reinforcement operations can be performed using a 3D model generated from multiple scans. For this purpose, an accurate alignment of the scanned models is required. Aligning existing structure model with data scanned after surface deformations can be challenging, particularly if reference markers are not available or were displaced. For scenarios where the surrounding structure is largely changed, certain procedures can be adapted when processing the scanned data to achieve consistent alignment between scanned and reference structure models. This work proposes a methodology to cope with these situations, analysing the impact of different approaches. Experiments were performed in a realistic scenario related with shotcrete of railway tunnels wall surfaces, with the results showing the applicability of the developed work. The proposed procedure relies in highlighting the importance of specific points that describe the same feature in the reference and aligning PC. The proposed methodology achieved an RMS difference of 0.0173 m, which lead to a drastic improvement in the point cloud alignment compared to the use of standard ICP algorithm without data preprocessing, which achieved 0.0518 m in the studied use-case.

Abstract
The advent of Industry 4.0 has created a need for more flexible and adaptable manufacturing systems. This paper proposes the integration of AAS (Asset Administration Shells), SBM (Skill-based manufacturing) and OPC UA (Open Platform Communications Unified Architecture), to enable more flexible manufacturing systems. The integration of these concepts provides a solution for achieving faster and easier dynamic reconfiguration in manufacturing systems, which is essential for fulfilling the demand of customization and flexibility in modern production systems. An Asset Administration Shell provides a standardized structure for describing assets and their administration, while Skill-based manufacturing enables the deployment of task-oriented machines that can self-configure, self-diagnose, and self-optimize their performance. The use of OPC UA as a communication protocol ensures that these systems can communicate with one another in a secure and reliable way. This paper presents a conceptual framework for the integration of these three open technologies. This framework contributes to having a single interface and source of information for every asset, which can lead to increased efficiency by reducing changeover times, thus reducing the overall cost in flexible manufacturing system scenarios. Future work will focus on the implementation and validation of this framework in a real-world manufacturing setting.