Application of Neural Radiance Fields for Autonomous Spacecraft Navigation and Planetary Characterization

With continuous lunar operations, small body sample returns, and distant planetary imaging missions slated for future NASA missions, advances in autonomous spacecraft navigation are necessary to enable onboard decision-making and reduce reliance on human and Deep Space Network (DSN) resources. Recent applications of deep learning to spacecraft pose estimation have demonstrated promising results for enabling greater autonomy and improved performance in extreme environments. Within this category of deep learning are neural radiance fields (NeRFs), which use multi-layer perceptron (MLP) neural networks to intake position and viewing direction inputs and produce color and volume density outputs. These outputs can be used to render a high fidelity 2D representation of a given scene, with demonstrated applications toward novel viewpoint and illumination synthesis. The proposed research presents three variations on the traditional NeRF architecture that can be applied toward extracting pose and reflectance model information as well as validating results generated by this approach. First, neural reflectance fields are defined as an expansion on the inputs and outputs of traditional NeRF, incorporating reflectance model parameters that are estimated through observations of the given scene. Second, an "inverted" NeRF (iNeRF) architecture is presented, which uses an initial pose guess and captured image to iteratively render a scene at from new perspectives until a sufficient pose estimate is achieved. Third, a verification and validation NeRF architecture offers a method for defining the system-level confidence in a given pose estimate based on a third rendered perspective. Cascading these three NeRF architectures together presents a fully autonomous, deep learning pose and reflectance estimation system that can perform tasks related to for on-orbit navigation, EDL, or surface operations.

This is early/mid-stage (TRL 2) — 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.