Completed two distinct projects in simulation: developing state estimation algorithms for an aerial drone and optimizing navigation stacks for a ground robot in complex environments.
Autonomous Robotics: Drone & Ground Systems
Objective
Key Contributions
- Drone State Estimation (C++): Implemented the behavior controller node and programmed an Extended Kalman Filter (EKF) from scratch to fuse IMU, magnetometer, GPS, and sonar data, enabling precise 3D localization.
- Robot Navigation (ROS 2): Tuned path planning algorithms (A*, Theta*) and costmap parameters to ensure collision-free navigation in tight corridors.
- Trajectory Control: Optimized the Regulated Pure Pursuit controller settings to eliminate oscillations, ensuring smooth path tracking for the ground vehicle.
Visuals
Fig 1. Path Planning Algorithm Comparison
Fig 2. A* Path Smoothing Analysis
Fig 1. Path Planning Algorithm Comparison
Fig 2. A* Path Smoothing Analysis