Kalman Filter & JS-Based Correction Mechanism

1. Kalman Filter development in JavaScript

- Implementation of the Kalman filter in JavaScript with functions for:
- Prediction: Calculation of the predicted state from current positions and speeds.
- Update: Incorporation of new sensor measurements (GPS, Steps, speed.) in the report.
- Delivery: A standalone JavaScript module (ES6) for direct integration into the project.

2. Implementation of the Correction Mechanism via OSRM in JavaScript
- Development of the map-matching logic with OSRM to align estimated positions on roads.
- Post-processing: Correction after each filter prediction.
- Real-time loopback: Integration of the corrected position as a measure for the Kalman filter.
- Delivery: JavaScript module connected to the OSRM API, integration into the processing pipeline.

3. Reference Trajectory Tests
- Selection of reference trajectories (car, pedestrian, dog).
- Performance and filter accuracy tests on these trajectories to validate the correct behaviour.
- Parameter adjustment to improve robustness and accuracy of estimates.

4. Drafting of a Technical Report
- Explanations of the functioning of the Kalman filter used.
- Scenario test results (accuracy of estimates, alignment with the road).
- Proposals for future improvements.