Lockheed Martin reported that its research collaboration with NVIDIA is demonstrating how Artificial Intelligence can combine data and detect anomalies in current environmental conditions. Their project was announced 14 months ago to create a digital twin for the National Oceanic and Atmospheric Administration (NOAA) to monitor and model global weather events and inform research into major scientific issues.
NVDIA is a hardware and software developer that specializes in data science and high-performance computing.
“A platform such as our digital twin could serve as a one-stop-shop for global weather monitoring and could enable more accurate initial conditions for predictive forecasting,” stated Lockheed AI research engineer Lynn Montgomery. “Every day, NOAA receives terabytes of weather data from numerous space and Earth-based sensors, and this project helps fuse this information accurately to support timely depiction of global conditions.”
The late 2022 contract from NOAA assigned Lockheed with developing the Earth Observations Digital Twin to monitor current global environmental conditions, including extreme weather events. The EODT is built on Lockheed’s OpenRosetta3D platform, with NVIDIA’s Omniverse Nucleus database engine, and provides 4D visualizations of the collected information.
Recently, the EODT demonstrated global sea-surface temperatures based on multi-sensor fusion of data from satellites and models, plus short term temperature anomalies.
According to Lockheed, in the months ahead the EODT will incorporate additional data streams, like space weather and sea ice concentrations.
NOAA’s aim is for the digital twin to provide an efficient and centralized approach to collect and combine, and then visualize data from various space and earth sensors. The sea-temperature demo showed NOAA and other government customers the potential of using AI to display high-resolution, accurate, and timely depictions of satellite and sensor data.
