← Back to NASA Technology Projects
Completed TRL 2 (started at 2, targeting 3)
The accuracy of future EDL missions will be improved by enhancing the atmospheric characterization process, and this can be used by future projects without additional development cost. All Mars EDL missions rely on flight mechanics simulations to predict performance, and these simulations are in turn reliant on quality atmospheric models that characterize the environment over challenging Martian terrains or variability with seasons like dust season. Atmosphere, especially winds, is a large component of EDL performance at Mars [Dutta 2017]. Current best practice in atmospheric modeling for Mars EDL simulations rely on mesoscale models [Mischna 2022, Dutta 2023], but the discretization of the data for simulations and characterizing bounds, like mean and standard deviation, does not maintain the physics. For example, spatial correlations might be broken to compute statistics from many different days of the data. Developing a Machine Learning (ML) or reduced order-based model will allow all future Mars missions, whether low-cost or flagship, to train models to the higher-fidelity mesoscale data and capture the inherent physics in the base atmospheric data while still generating dispersed atmospheric profiles for EDL simulations.
Current atmosphere models for EDL employ simplistic but effective pre-processing of the mesoscale atmosphere to produce statistical metrics, like mean and standard deviation, and allow flight simulations to disperse atmospheric conditions for statistical simulations that create performance metrics used by mission designers. Advances in Artificial Intelligence (AI)/ML models and other techniques like reduced order modeling can capture the atmospheric related physics inherent in the raw mesoscale data, but once the model is trained, may perturbed profiles of atmospheric properties can be created for the flight mechanics simulations without sacrificing the atmospheric physics. Although ML and reduced order model for one set of training data (and scenario) are not transferrable to another scenario, the inherent process can be used for many different Mars applications and locations. Also, once a model is trained for a given scenario, it can generate as many dispersed atmosphere that is needed by the simulation. Such ML and reduced order models have not been applied to atmospheric modeling for simulations in the past and the overall cost to a future Mars missions is small once the process has been developed. This process can provide a cheap way to generate dispersed atmospheric profiles.
Listed on TechPort itself — the most direct way to ask about this specific project.
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.