Saraiva, António Pedro e Pedro2026-01-152026-01-152025http://hdl.handle.net/10400.5/116642Tese de Mestrado, Engenharia Informática, 2025, Universidade de Lisboa, Faculdade de CiênciasIn recent years, legged robots have emerged as a promising solutionto navigate complex and unstructured environments, combining the advantages of mobility and adaptability. This thesis presents a comprehensive study conducted in two distinct phases, focusing on the development and implementation of autonomous navigation systems for quadrupedal robots in unknown environments. The first phase covers the complete construction and integration of a quadrupedal robot using the hands on robotics.org kit, providing valuable information on the mechanical and electrical systems that govern legged locomotion. Thishands-on approach established a fundamental understanding of the challenges and constraints in herent to quadrupedal platforms. The second phase, conducted during an Erasmus exchange program, uses the advanced capabilities of the A1 Unitree robot to implement and validate sophisticated mapping and path planning algorithms. This system enables the robot to autonomously generate environmental maps and compute optimal trajectories using A* search toward user-defined goals, effectively addressing the challenges of unknown environment navigation. The research contributes to the field by demonstrating the practical implementation of autonomous navigation systems on different quadrupedal platforms, having a 70% success rate in complex maze environments with the A1 robot. The results show the effectiveness of the proposed approach in generating efficient paths while maintaining stable locomotion under various environmental conditions. This work provides valuable insights for the future development of autonomous legged robots capable of operatingin real-world scenarios.application/pdfengQuadrupedal RoboticsAutonomous NavigationPath PlanningEnvironmental MappingRobot Locomotionmaster thesis204176859