SBIR-STTR Award

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project is that it will provide a concrete implementation with practical commercial applications in renewable energy and climate-related risk of a hybrid, ultrafast physics-informed machine learning technology that emulates complex numerical physics-based climate/weather models. Physics-based (hydro)climate/weather simulation models are used across trillion-dollar industries of utmost societal interest, from agriculture to insurance to energy to logistics. Faster (by 3-5 orders of magnitude), hyperlocal, large-scale estimates of physical climate/environmental parameters that are difficult/expensive or even impossible to measure empirically (such as snow-water equivalent), integrating best-available real-time observational remote-sensing data, can both streamline existing applications (faster hydropower scenario forecasting), as well as enable new capabilities and products (e.g., real-time storm risk response or automated parametric insurance contracts). The proposed R&D effort will illustrate how scientific modeling, including of climate, can leverage both the body of knowledge embedded in numerical simulation models, which the scientific community has spent more than seven decades building, as well as the high speed and natural capability of novel AI and machine learning models to process novel sources of observational data (particularly remote-sensing) on the natural environment. This Small Business Innovation Research (SBIR) Phase I project addresses the need in the renewable energy and insurance industries for fast, high-resolution (in space and time) estimates of the hazard profiles of environmental and climate/weather parameters informed by real-time observational data. The project aims to provide a first proof-of-concept that a commercial-grade hybrid physics-informed AI technology can be developed for estimating relevant climate and weather parameters, starting with hydroclimate modeling. The R&D effort proposed will focus on 1) developing and validating a generative deep learning model trained on numerical hydroclimate simulation data as well as observational meteorological data; 2) identifying and benchmarking best-practices for ensuring stable training and updating of the model, observational/simulation data requirements, and computational resources needed; and 3) designing and developing streamlined model access patterns and web-based API functionality for use cases relevant to renewable energy and insurance/risk modeling use-cases. The envisioned proof-of-concept is a modular computational system running natively on GPU hardware that will allow creating gridded datasets of physical parameters such as snow water equivalent, precipitation, or water level, as well as their associated probability curves for geographical locations and time horizons of interest. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

The broader impact of this Small Business Innovation Research (SBIR) Phase II project is to provide commercially-deployable technology for highly-scalable, spatially-granular, and cost-effective risk predictions of climate-driven events, such as wildfire spread, from real-time to yearly time scales. As insured losses due to wildfires have increased over fivefold in the last decade, the associated risk makes it critical to improve the ability to predict physical and financial impacts at scale. Current predictive technologies used in major industries, like energy and insurance, are based on complex, hand-engineered, and computationally-intensive numerical physics models of climate and weather. In contrast, the proposed technology develops special AI emulator systems that learn the relevant physics and key drivers, including wind and surface hydrology, in wildfires. The proposed system can perform predictions much more efficiently due to a far simpler computational workflow and native AI hardware acceleration. In addition, AI emulators automate the assimilation of vastly higher amounts of remote-sensing and other observational data (e.g., radar measurements from weather satellites or land cover and vegetation data) over numerical models, allowing for increased accuracy, continuous improvement, and dynamic predictions reflecting changing on-the-ground conditions. This Small Business Innovation Research (SBIR) Phase II project addresses the pressing need in the energy and insurance industries to accurately and consistently assess wildfire risk over large geographical regions and at a localized level, on time scales ranging from daily to yearly. The proposed R&D will focus on developing and validating an AI emulator of wildfire spread. This entails 1) developing AI architectures for assimilating observational (remote-sensing) and numerical simulation data on drivers of wildfire at different temporal and spatial scales, including vegetation, soil hydrology, and atmospheric winds; 2) integrating data on historical wildfires and their spread to drive the learning process; 3) conducting extensive verification and validation studies; and 4) developing and deploying APIs and graphical interfaces for accessing AI emulator output on the cloud.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.