Abstract
The advancement of technology has led to a growing demand for autonomy across various sectors. A key aspect of achieving autonomous navigation through intricate environments is path planning, initially confined to 2D spaces but rapidly evolving to address the complexities of 3D environments. Despite the widespread adoption of RRT-based planners, their inherent lack of optimality has encouraged researchers to find refinements. This paper transposes an existing algorithm developed for 2D environments to 3D, leveraging a heuristic to optimize the generated paths in terms of path length, memory consumed, and execution time. Along with this scalability to 3D scenarios, a modification was introduced that trades off some execution time for a substantial improvement in path length. The results obtained from a series of simulated experimental tests prove the efficacy of the proposed method in 3D environments, demonstrating reduced memory consumption and execution time compared to conventional approaches.

Abstract
Numerous pose estimation methodologies demonstrate a decrement in accuracy or efficiency metrics when subjected to highly cluttered scenarios. Currently, companies expect high-efficiency robotic systems to close the gap between humans and machines, especially in logistic operations, which is highlighted by the requirement to execute operations, such as navigation, perception and picking. To mitigate this issue, the majority of strategies augment the quantity of detected and matched features. However, in this paper, it is proposed a system which adopts an inverse strategy, for instance, it reduces the types of features detected to enhance efficiency. Upon detecting 2D polygons, this solution perceives objects, identifies their corners and edges, and establishes a relationship between the features extracted from the perceived object and the known object model. Subsequently, this relationship is used to devise a weighting system capable of predicting an optimal final pose estimation. Moreover, it has been demonstrated that this solution applies to different objects in real scenarios, such as intralogistic, and industrial, provided there is prior knowledge of the object's shape and measurements. Lastly, the proposed method was evaluated and found to achieve an average overlap rate of 89.77% and an average process time of 0.0398 seconds per object pose estimation.

Abstract
The high level of precision and consistency required for pallet detection in industrial environments and logistics tasks is a critical challenge that has been the subject of extensive research. This paper proposes a system for detecting pallets and its pockets using the You Only Look Once (YOLO) v8 Open Neural Network Exchange (ONNX) model, followed by the segmentation of the pallet surface. On the basis of the system a pipeline built on the ROS Action Server whose structure promotes modularity and ease of implementation of heuristics. Additionally, is presented a comparison between the YOLOv5 and YOLOv8 models in the detection task, trained with a customised dataset from a factory environment. The results demonstrate that the pipeline can consistently perform pallet and pocket detection, even when tested in the laboratory and with successive 3D pallet segmentation. When comparing the models, YOLOv8 achieved higher average metric values, with YOLOv8m providing better detection performance in the laboratory setting.

Abstract
This paper discusses the emerging field of robotics, particularly focusing on motion planning for robotic manipulators. It highlights the need for simplification and standardization in robot implementation processes. Among several tools available, the paper focuses on the MoveIt tool due to its compatibility, popularity, and community contributions. However, the paper acknowledges some resistance in developing new applications with MoveIt, especially for researchers and beginners. To address this, the paper introduces an efficient, modular action server for interacting with the MoveIt framework. This pipeline simplifies parameter reconfiguration and provides a general solution for the motion planning problem. It can calculate trajectories for robotic manipulators without environmental collisions using a single server request and supports operation in different modes. The server was tested on an Universal Robots UR10 manipulator, demonstrating its ability to quickly plan paths for two test operations: an object pick-and-place mission and a collision avoidance test. The results were positive, achieving the set goals with minimal user-server interaction. This work represents a significant step towards more efficient and user-friendly robotic manipulation.

Abstract
Bin picking is a challenging task involving many research domains within the perception and grasping fields, for which there are no perfect and reliable solutions available that are applicable to a wide range of unstructured and cluttered environments present in industrial factories and logistics centers. This paper contributes with research on the topic of object segmentation in cluttered scenarios, independent of previous object shape knowledge, for textured and textureless objects. In addition, it addresses the demand for extended datasets in deep learning tasks with realistic data. We propose a solution using a Mask R-CNN for 2D object segmentation, trained with real data acquired from a RGB-D sensor and synthetic data generated in Blender, combined with 3D point-cloud segmentation to extract a segmented point cloud belonging to a single object from the bin. Next, it is employed a re-configurable pipeline for 6-DoF object pose estimation, followed by a grasp planner to select a feasible grasp pose. The experimental results show that the object segmentation approach is efficient and accurate in cluttered scenarios with several occlusions. The neural network model was trained with both real and simulated data, enhancing the success rate from the previous classical segmentation, displaying an overall grasping success rate of 87.5%.