Develop state estimation achritecture

We need to design and implement a state estimation architecture for the F1TENTH vehicle. This system will fuse data from various onboard sensors (e.g., odometry, IMU, LiDAR) to produce a reliable and consistent estimate of the vehicle's pose and velocity in real time. The estimated state will serve as a critical input to control, planning, and decision-making modules.

---

🎯 Goals

• Define the state variables to be estimated (e.g., position, orientation, linear and angular velocity)

• Identify available sensor inputs (e.g., wheel encoders, IMU, LiDAR, camera)

• Choose and document an appropriate estimation algorithm (e.g., Extended Kalman Filter, Unscented Kalman Filter, factor graph-based SLAM, etc.)

• Design a modular state estimation pipeline that:

  • Handles sensor data streams asynchronously
  • Applies sensor fusion to improve robustness
  • Outputs estimated state at a fixed rate

• Define input/output interfaces for integration with other modules

---

📝 Notes

• The estimator should run in real-time and be resilient to occasional sensor dropout or noise.
• Consider open-source packages (e.g., robot_localization, ekf_localization, or custom implementations) and document pros/cons.
• Ensure the system is extensible—e.g., able to incorporate GPS or visual odometry in the future.
• Logs or visualization tools (e.g., RViZ, rqt_plot) should be used for debugging and validation.

---

📂 Deliverables (optional but encouraged)

• Diagram of the state estimation architecture
• Configurable YAML or launch files (if using ROS-based tools)
• Sample plots of estimated vs. ground truth state (if available)