I'm currently pursuing a degree in Mechatronics Engineering at the São Carlos School of Engineering (EESC-USP). I'm passionate about technology, with a strong interest in robotics and artificial intelligence.
I have experience with various programming languages and am always eager to learn new technologies and tools. My current focus is on projects related to robotics and machine learning, where I aim to push the boundaries of innovation.
Robotic manipulator designed for thinning and harvesting table grapes. Developed as an academic project to design and build a functional manipulator.
-
graspe-v3 (newer)
- Real robot, 4 DoFs + 1 gripper
- PID control implementation
- Serial communication with a computer
- Interactive GUI
- Simulation using the Robotics Toolbox (Peter Corke)
-
graspe (v1 and v2, deprecated)
-
v1: Simulated robot using ROS 2 and ros_control
- ROS 2 in a Docker environment
- Implementation of a simple trajectory using ROS 2 + ros_control
- Simulation in Gazebo
-
v2: Real robot, using servos
- Servo motors with factory controllers
- Xbox controller interface in the workspace
- New mechanical design
-
Research project at LabRom (EESC-USP) involving:
- Deep RL for obstacle avoidance
- High-level policy predicting velocity commands from LiDAR, IMU, and odometry
- Low-level controller for joint-level tracking
- Experiments on the Unitree Go1, including comparisons with classical methods (MPC + TEB)
- Garcia, J. P. B. et al., Neural Network-Based Velocity Control for Dynamic Obstacle Avoidance in Legged Robots — 1st Brazilian Conference on Robotics - CROS (2025).