Robotics & simulation
SLAM & Kalman Localization
Making vehicle localization more reliable when sensor readings are noisy.
Conceptual architecture illustration
The idea
A vehicle navigation project in ROS2 and Webots, using a simulated Tesla Model 3 to explore mapping, localization, and obstacle avoidance. An Extended Kalman Filter brings sensor readings together to address drift and positioning uncertainty.
Inside the build
- 01
Implemented an Extended Kalman Filter for vehicle localization.
- 02
Combined LiDAR and camera data for environmental perception.
- 03
Used simulation to iterate on lane following and obstacle avoidance.