Sensor Fusion & State Estimation
Design and tune KF/EKF/UKF/ESKF sensor fusion, and prove the filter is consistent with NEES/NIS tests.
$39
Works with the AI Agents you already use
Sensor Fusion & State Estimation
Example session with this skill installed
I'm seeing jumps in my drone's position when fusing GNSS and IMU. Help me set up an EKF state vector and a validation plan to check for filter consistency.
- Read your context and instructions
- Compiled the sensor fusion
- Generated the data export
I have defined the kinematic state vector and the corresponding validation workflow to address the position jumps. The focus is on ensuring the attitude and position estimates remain consistent with the sensor noise profiles through NEES/NIS testing.
EKF Configuration and Validation Plan
| Component | Definition / Specification |
|---|---|
| State Vector ($x$) | $[p_n, p_e, p_d, v_n, v_e, v_d, q_w, q_x, q_y, q_z, b_{acc}, b_{gyro}]^T$ |
| Process Model | IMU-driven kinematics (15-state) with bias random walk integration. |
| Measurement 1 | GNSS Position ($p_n, p_e, p_d$) with innovation gating. |
| Measurement 2 | IMU Attitude / Angular Velocity for prediction step. |
| Validation Script | python scripts/nees_check.py errors.csv |
| Consistency Metric | Normalized Innovation Squared (NIS) < $\chi^2$ threshold (0.95 confidence). |
Validation Steps
Static Log Analysis: Record 2 hours of static IMU data to calculate Allan Variance for noise densities ($N$) and random walks ($K$).
2.
Jacobian Verification: Ensure the measurement model $H$ for GNSS is correctly linearized at the current state estimate.
3.
Innovation Gating: Apply a Mahalanobis distance check on GNSS updates to reject multipath outliers that cause position jumps.
4.
NEES Test: Run the filter against ground truth data in simulation to ensure the error stays within the 95% confidence interval of the covariance $P$.
Next steps
- Provide a sample of your static IMU log to calculate the noise parameters.
- Review the
nees_check.pyrequirements to ensure your error logs are formatted correctly. - Implement the innovation gate threshold (typically set to 5.99 for 2-DOF position updates).
errors.csv
CSV · data export
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Know whether your filter is lying — before the robot does.
Design, implement and validate state estimation by sensor fusion - IMU, GNSS/RTK, wheel odometry, lidar, camera, barometer, magnetometer - with Kalman filters (KF/EKF/UKF/ESKF), particle filters and factor graphs, including noise tuning from data and consistency tests (NEES/NIS) using a tested checker. Use when the user fuses sensors for a robot, drone or vehicle, sees drift or jumps in position/attitude, or needs to tune covariances.
Includes an EXPERT.md reference (exact parameters, thresholds, failure diagnosis). Tested scripts: scripts/nees_check.py.
Acceptable use: civil and commercial applications only. Any military, paramilitary or weapon-related use is prohibited, as stated in the skill itself.
Restricted to civil/commercial use. Refuses military, weapon, or unauthorized surveillance applications. Requires user-provided sensor logs for noise tuning.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 28 days ago