Multi-path Fusion Machine Learning for New Observing System Design and Operations

We propose to develop a system based on data fusion and multi-path neural network Machine Learning (ML) to aid in the design and operation of multi-sensor New Observing Systems (NOS). This involves building ML-enabled analytic tools and advanced computing environment capabilities for the creation of a NOS workflow that utilizes large amounts of diverse airborne and satellite observations. Rather than a ML model consisting of a single stack of layers with a single input, we will have multiple neural networks working in parallel. During development we will demonstrate the capabilities using a forest productivity use case, in which fusion of lidar, spectrometry, satellite-derived climatology and ecosystem modeling will provide insights into the driving environmental factors that influence productivity. Our concept uses terrestrially based, maximum-accuracy models for NOS design purposes, and will use them to conduct sensitivity studies to guide sensor and mission requirements traceability. Additionally, we will benchmark on-orbit, light-weight models based on these terrestrial model on flight-like hardware in order to enable onboard decision making such as intelligent targeting and compressive sensing.

Advance Earth system science knowledge through the Identification, develop, and demonstrate innovative information systems technologies

This is early/mid-stage (TRL 5) — 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.