Improved Autonomous Navigation Through Optimal Sensor Outliers

One of the greatest challenges associated with pose estimation using cameras or sensors such as LIDAR is removal of sensor outliers. Outliers exist for numerous reasons, such as incorrectly associated feature points, extraneous data points, etc. Without proper removal of outliers, an incorrect pose estimate may result, which can lead to difficulty in achieving mission goals for autonomous operations. A common approach to remove outliers is the RANSAC algorithm. But RANSAC may not be optimal, especially if an incorrect model is used, thus leading to outliers that may pass the RANSAC test but may degrade the pose estimate. In Phase I an optimal closed-form approach to replace the RANSAC algorithm was shown that it is based on using the statistical properties of the sensor error in its derivation. The heart of the solution is based on an optimally derived pose estimation solution from a Total Least Squares (TLS) approach. A byproduct of the TLS solution is the covariance of the sensor residuals, which is the key to remove outliers. The Phase I work showed that because the covariance is optimal, it results in a more robust approach to remove outliers than the standard RANSAC algorithm and its variants. The Phase I work focused on studying the effectiveness of the newly derived closed-form covariance in a new RANSAC-type algorithm, called the Statistical Optimal RANSAC (SO-RANSAC) algorithm. For the Phase II work, SO-RANSAC will be further refined to include the two-camera scenario. Invariant methods will also be employed to initialize the SO-RANSAC algorithm. Also, a colored-noise filter will be implemented to handle heavy-tail errors associated with feature mis-associations. Furthermore, studies will be performed to provide the computational effort required to implement SO-RANSAC on a real spacecraft processor. All these efforts will significantly increase the TRL capability of SO-RANSAC, leading to an actual implementation capability. The significance of proposed technology is the development of an optimal feature outlier detection algorithm that can be used in a navigation system for an arbitrary noncooperative object.  The developed algorithm consists of extracting feature information from images collected by the optical sensor or point clouds from LIDAR-type measurements.  The new algorithm, called Statistical Optimal RANSAC (SO-RANSAC), is based on a rigorously derived solution using the statistical properties of the sensor errors.  This new algorithm will ensure that the features are accurately extracted, thus providing a robust approach compared to the current state-of-the-art, such as traditional RANSAC algorithms. Technical Objectives Objective 1 − Expand the Theoretical Developments to Include the Two-Camera Case Objective 2 − Implement an Invariant-Based Outlier Rejection for Initialization Objective 3 − Assess the Computational Effort Versus the Performance   Work Plan Summary The Phase II effort will focus on relative spacecraft navigation algorithm development and extensive simulation studies.  This will build a foundation to develop benchmark testing at the onset of Phase III, with the end of this work being a fully functional demonstration unit.  This can lead to a future mission to test the proposed technology.   Proposed Deliverables Bimonthly Progress Reports, Prototype Software, Simulation Studies and a Final Report.

NASA has flown several formation flying missions, such A-Train and Cluster.  Also, applications involving proximity operations are of great interest to NASA, as well as safe, precision landing on small bodies.  The proposed technology further advances current navigation applications related to all these applications since it provides a robust solution for noncooperative objects.  The application is based on rigorously derived error definitions, so that physically correct uncertainty bounds are provided for feature outlier mitigation. Non-NASA applications, especially DoD ones, are heavily focusing on anti-jamming communication and navigation systems, such as GPS-less navigation.  The proposed technology can significantly advance these focus areas because it is self-contained and decreases the susceptibility to outside attacks.

This is early/mid-stage (TRL 3) — the most realistic path in is NASA SBIR/STTR, which funds small businesses and research institutions to develop technology aligned with NASA's needs (equity-free, phased funding). Check whether a current SBIR/STTR solicitation topic overlaps with this project's technology area, or contact the project directly (above) to ask.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.