Abstract
The robot presented in this paper was developed with the main focus on participating in robotic competitions. Therefore, the subsystems here presented were developed taking into account performance criteria instead of simplicity. Nonetheless, this paper also presents background knowledge in some basic concepts regarding robot localization, navigation, color identification and control, all of which are key for a more competitive robot.

Abstract
Object localization and tracking is core to many practical applications, like human-computer interaction, security and surveillance, robot competitions and Industry 4.0. Such task may be computationally demanding, especially for traditional embedded systems, that usually have tight processing and storage constraints. This calls for the investigation of alternatives, including emergent heterogeneous embedded systems, like the Parallella line of single-board-computers (SBCs). The work presented in this paper explores the use of a Parallella board with a 16-core Epiphany co-processor, to perform real-time tracking of objects in frames captured by a Kinect sensor, based on color segmentation. We addressed several processing strategies, trying to assess which one performed better. We also ran the same code (where applicable) in several models of the Raspberry Pi platform, for comparison. We conclude that effectively exploring the Epiphany co-processor is not trivial, requiring considerable programming effort and suitable applications (CPU-demanding and highly parallelizable), to the extent that simpler development approaches, on more recent SBCs may be more effective. © 2017.

Abstract
The use of machine learning in embedded systems is an interesting topic, especially with the growth in popularity of the Internet of Things (IoT). The capacity of a system, such as a robot, to self-localize, is a fundamental skill for its navigation and decision-making processes. This work focuses on the feasibility of using machine learning in a Raspberry Pi 4 Model B, solving the localization problem using images and fiducial markers (ArUco markers) in the context of the RobotAtFactory 4.0 competition. The approaches were validated using a realistically simulated scenario. Three algorithms were tested, and all were shown to be a good solution for a limited amount of data. Results also show that when the amount of data grows, only Multi-Layer Perception (MLP) is feasible for the embedded application due to the required training time and the resulting size of the model.

Abstract
Self-localization of a robot is one of the most important requirements in mobile robotics. There are several approaches to providing localization data. The Ultra Wide Band Time of Flight provides position information but lacks the angle. Odometry data can be combined by using a data fusion algorithm. This paper addresses the application of data fusion algorithms based on odometry and Ultra Wide Band Time of Flight positioning using a Kalman filter that allows performing the data fusion task which outputs the position and orientation of the robot. The proposed solution, validated in a real developed platform can be applied in service and industrial robots.

Abstract
Indoor Air Quality (IAQ) pertains to the air quality within a specific space and is directly linked to the well-being and comfort of its occupants. In line with this objective, this research presents a real-time system dedicated to monitoring and predicting IAQ, encompassing both thermal comfort and gas concentration. The system initiates with a data acquisition, wherein a set of sensors captures environmental parameters and transmits this data for storage in a database. The measured parameters are analyzed by a neural network algorithm that predicts anomalies based on historical data. The neural network model generated predictions from 75.9% to 98.1% (depending on the parameter) of precision during regular situations. After that, a test with smoke in the same place was done to validate the model, and the results showed it could detect anomalies. Finally, prediction data are stored in a new database and displayed on a dashboard for monitoring in real-time measured and prediction data. © 2023 IEEE.

Abstract
The accurate measurement of joint angles during patient rehabilitation is crucial for informed decision making by physiotherapists. Presently, visual inspection stands as one of the prevalent methods for angle assessment. Although it could appear the most straightforward way to assess the angles, it presents a problem related to the high susceptibility to error in the angle estimation. In light of this, this study investigates the possibility of using a new approach to angle calculation: a hybrid approach leveraging both a camera and LiDAR technology, merging image data with point cloud information. This method employs AI-driven techniques to identify the individual and their joints, utilizing the cloud-point data for angle computation. The tests, considering different exercises with different perspectives and distances, showed a slight improvement compared to using YOLO v7 for angle calculation. However, the improvement comes with higher system costs when compared with other image-based approaches due to the necessity of equipment such as LiDAR and a loss of fluidity during the exercise performance. Therefore, the cost-benefit of the proposed approach could be questionable. Nonetheless, the results hint at a promising field for further exploration and the potential viability of using the proposed methodology.