Abstract
Purpose: Visual perception enables robots to perceive the environment. Visual data is processed using computer vision algorithms that are usually time-expensive and require powerful devices to process the visual data in real-time, which is unfeasible for open-field robots with limited energy. This work benchmarks the performance of different heterogeneous platforms for object detection in real-time. This research benchmarks three architectures: embedded GPU-Graphical Processing Units (such as NVIDIA Jetson Nano 2 GB and 4 GB, and NVIDIA Jetson TX2), TPU-Tensor Processing Unit (such as Coral Dev Board TPU), and DPU-Deep Learning Processor Unit (such as in AMD-Xilinx ZCU104 Development Board, and AMD-Xilinx Kria KV260 Starter Kit). Methods: The authors used the RetinaNet ResNet-50 fine-tuned using the natural VineSet dataset. After the trained model was converted and compiled for target-specific hardware formats to improve the execution efficiency.Conclusions and Results: The platforms were assessed in terms of performance of the evaluation metrics and efficiency (time of inference). Graphical Processing Units (GPUs) were the slowest devices, running at 3 FPS to 5 FPS, and Field Programmable Gate Arrays (FPGAs) were the fastest devices, running at 14 FPS to 25 FPS. The efficiency of the Tensor Processing Unit (TPU) is irrelevant and similar to NVIDIA Jetson TX2. TPU and GPU are the most power-efficient, consuming about 5 W. The performance differences, in the evaluation metrics, across devices are irrelevant and have an F1 of about 70 % and mean Average Precision (mAP) of about 60 %.

Abstract
The efficient application of phytochemical products in agriculture is a complex issue that demands optimised sprayers and variable rate technologies, which rely on advanced sensing systems to address challenges such as overdosage and product losses. This work developed a system capable of processing different tree canopy parameters to support precision fruit farming and environmental protection using intelligent spraying methodologies. This system is based on a 2D light detection and ranging (LiDAR) sensor and a Global Navigation Satellite System (GNSS) receiver integrated into a sprayer driven by a tractor. The algorithm detects the canopy boundaries, allowing spray only in the presence of vegetation. The spray volume spared evaluates the system's performance compared to a Tree Row Volume (TRV) methodology. The results showed a 28% reduction in the overdosage of spraying product. The second step in this work was calculating and adjusting the amount of liquid to apply based on the tree volume. Considering this parameter, the saving obtained had an average value for the right and left rows of 78%. The volume of the trees was also monitored in a georeferenced manner with the creation of a occupation grid map. This map recorded the trajectory of the sprayer and the detected trees according to their volume.

Abstract
[No abstract available]

Abstract
Digital twin is one promising and key technology that emerged with Industry 4.0 to assist the decision-making process in multiple industries, enabling potential benefits such as reducing costs, and risk, improving efficiency, and supporting decision-making. Despite these, the decision-making approach of carrying out a what-if simulation study using digital twin models of each and every possible scenario independently is time-consuming and requires significant computational resources. The integration of recommendation systems within the digital twindriven decision-support framework can support the decision-making process by providing targeted scenario recommendations, reducing the decision-making time and imposing decision- making efficiency. However, recommendation systems have inherent challenges, such as cold-start, data sparsity, and prediction accuracy. The integration of trust and similarity measures with recommendation systems alleviates the challenges mentioned earlier, and the integration of machine learning techniques enables better recommendations through their ability to simulate human learning. Having this in mind, this paper proposes a trust-based recommendation approach using a reinforcement learning technique combined with similarity measures, which can be integrated within a digital twin-based what-if simulation decision-support system. This approach was experimentally validated by performing accurate recommendations in an industrial case study of a battery pack assembly line. The results show improvements in the proposed model regarding the accuracy of the prediction about the user rating of the recommended scenarios over the state-of-the-art recommendation approaches, particularly in coldstart and data sparsity scenarios.

Abstract
Long-term operation of robots creates new challenges to Simultaneous Localization and Mapping (SLAM) algorithms. Long-term SLAM algorithms should adapt to recent changes while preserving older states, when dealing with appearance variations (lighting, daytime, weather, or seasonal) or environment reconfiguration. When also operating robots for long periods and trajectory lengths, the map should readjust to environment changes but not grow indefinitely. The map size should depend only on updating the map with new information of interest, not on the operation time or trajectory length. Although several studies in the literature review SLAM algorithms, none of the studies focus on the challenges associated to lifelong SLAM. Thus, this paper presents a systematic literature review on long-term localization and mapping following the Preferred Reporting Items for Systematic reviews and Meta-Analysis guidelines. The review analyzes 142 works covering appearance invariance, modeling the environment dynamics, map size management, multisession, and computational topics such as parallel computing and timing efficiency. The analysis also focus on the experimental data and evaluation metrics commonly used to assess long-term autonomy. Moreover, an overview over the bibliographic data of the 142 records provides analysis in terms of keywords and authorship co-occurrence to identify the terms more used in long-term SLAM and research networks between authors, respectively. Future studies can update this paper thanks to the systematic methodology presented in the review and the public GitHub repository with all the documentation and scripts used during the review process.

Abstract
The pose estimation of a mobile robotic system is essential in many autonomous applications. Inertial sensors provide high-frequency measurements that can be used to estimate the displacement, however, for estimating the orientation, an additional filter is required. Some of the newest Attitude and Heading Reference Systems can provide a referenced estimation of the orientation of the device, allowing it to retrieve the orientation of a robotic system. However, magnetic field perturbations caused by ferromagnetic objects or induced magnetic fields might influence these systems and, consequently, lead to the accumulation of errors over time. In this paper, the performance of the Xsens fusion filter is compared with a stateof-the-art algorithm to estimate the orientation of the system under dynamic movements and in the presence of magnetic perturbations, with the goal of finding the most suitable for an Unmanned Aerial Vehicle. The results show that both filters are robust and perform well in the target scenario, with a root mean squared error between 2 and 5 degrees; however, the Xsens fusion filter does not require an extra computer to process the data.