Development and implementation of an open source quadruped robot

This work presents an open-source quadruped robot, Naranjelio, featuring a 3Dprinted structure, cost-effective servo motors, and an onboard IMU and Raspberry Pi for real-time feedback. A ROS2-based control environment enables motion generation, data logging, and user interaction, while an Isaac Sim integration supports reinforcement learning exploration. A key contribution is the Quadruped Gait (E)Stability (QGE) metric, which combines acceleration and angular velocity data to yield a 0-1 gait quality score. Experimental results demonstrate Naranjelio's ability to carry payloads up to 66% of its own mass, though with reduced range of motion. Crucially, QGE quantifies these performance trade-offs, guiding iterative refinements in both structure and control for more robust, reliable gait.

• Autores: Correal Murillo, Nicolas
• Tipo de recurso: Trabajo de grado de pregrado
• Fecha de publicación: 2025
• Institución: Universidad de los Andes
• Repositorio: Séneca: repositorio Uniandes
• Idioma: eng
• OAI Identifier: oai:repositorio.uniandes.edu.co:1992/75538
• Acceso en línea: https://hdl.handle.net/1992/75538
• Palabra clave: Robótica; Analisis de movimiento; Diseño robotico; Cuadrupedo; Aprendizaje por refuerzo; Robotics; Movement analysis; Robot design; Quadruped; Reinforcement learning; Isaac Sim; ROS2; Ingeniería
• Rights: openAccess
• License: Attribution-NonCommercial-ShareAlike 4.0 International