Calibration with Machine Learning

Machine Learning (ML) has become a popular method to do model based computer processing without having to explicitly develop a model. Computer processing can be used for many things including calibration. Aerospace instrument design typically includes detectors that require calibration. Calibration for optical detectors can be quite extensive requiring time consuming, resource rich lab testing. For networks of instruments taking the same measurement (including space based sensor constellations), time and effort spent calibrating and maintaining calibration in flight for those instruments can be non-trivial. Reducing that time through the use of ML is the intent of this proposal. The idea is to learn (through ML) the characteristics of one reference instrument to then apply what is learned directly to any new (uncalibrated) instrument (of the same type) without going through the arduous calibrating process.

We believe the results from this work will mostly benefit low Size, Weight, and Power (SWAP) sensors/instruments where multiple instruments are deployed. Not only do we believe the calibration workflow for any given instrument can be reduced with this work, re-calibration should become easier.

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.