Curating Uncertainty for Reliable Exploitation and Collaboration (CURE-C)

Incomplete, noisy, or otherwise uncertain data is the way it is, no purely computational method can magically fix it. Numerous techniques developed by the database and programming languages communities exist to help users cope with uncertain data. Specifically, these tools allow users to define business logic over uncertain (i.e., incomplete, possibilistic, or probabilistic) data as if it were idealized deterministic data. The tool instruments the business logic (e.g., a relational database query) according to a model of uncertainty in the source data, to derive a model of uncertainty over the outputs. Tools for uncertain data management are crucial for working with real world data, but suffer from a range of usability challenges that this proposal aims to address. Curating Uncertainty for Reliable Exploitation and Collaboration (CURE-C), developed by XAnalytix Systems and University at Buffalo will solve this problem by building around a collaborative data store that manages both raw and curated data. Sensor data is ingested into the data store, where individuals that we refer to as Curators examine the data, address problems with it, and develop uncertainty profiles for the data. These efforts, once completed are recorded into a workflow and used to process new data as it arrives. These workflows apply the uncertainty profiles to the data, resulting in uncertainty-annotated data. Uncertainty profiles developed by the curators are also saved in a profile store to assist in future curation efforts. Planners can then pose queries over the uncertainty-annotated data, and receive results with both quantitative and qualitative uncertainty descriptions. At the completion of Phase I CURE-C will have a set of interfaces that streamline the development of uncertainty profiles by facilitating the re-use of previously developed profiles. CURE-C will also have an integrated analysis of multiple classes of uncertainty.

NASA is currently using AI-based techniques for many applications, including: identifying nearby asteroids, finding exoplanets, and predicting extreme solar radiation events. The Frontier Development Lab (FDL) is partnership with NASA and the DOE in the US and ESA in Europe. Each year it puts out a challenge short list that involves AI techniques. A past challenge involved testing the capability of a Bayesian neural network against a widely used machine learning techniques, which requires conditional probabilities that are accurate.

Intelligent manufacturing integrates information technology and manufacturing technology through a human-cyber-physical system (HCPS). ” The “Fourth Industrial Revolution” (or Industry 4.0) is the next evolutionary stage of current industrial manufacturing processes. This is a vast market opportunity for aspects that involve uncertainty between automated machines and humans.

This is early/mid-stage (TRL 4) — 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.