Lockheed Martin Tracks Wildland Fires with Cesium
It’s wildland fire season in the United States. The Lockheed Martin AI Center’s (LAIC) Joint Activity Manager (JAM) provides operational context for active wildland fires, using AI, real-time data, and CesiumJS to stream this 3D visualization to incident commanders.
Overview of Lockheed Martin JAM capability, featuring a sped-up clip of the Vista fire outside of Los Angeles, California. Courtesy Lockheed Martin.
While still in development, this capability is intended for state agencies to visualize and analyze active wildland fires; track flights of aerial assets that have been tasked with detection, monitoring, and suppression; and incorporate weather information, such as temperature, humidity, and wind.
The wildland fire incident list in the application is populated every 24 hours with information from the National Interagency Fire Center (NIFC), based in Boise, Idaho, USA, to ensure daily updates on the number and locations of active wildland fires. Imagery from National Oceanic and Atmospheric Administration’s (NOAA) Geostationary Operational Environmental Satellites (GOES) is pulled every 30 minutes.
In an upcoming feature, Lockheed Martin’s AI will detect near-real-time fire hot spots from this imagery and reflect them atop 3D terrain from Cesium every 15-20 minutes. Lockheed Martin uses its “super resolution” AI capabilities to, in effect, add more pixels to the images, which refines them prior to determining the near-real-time fire area from the satellite imagery.
Tracking a Pilatus PC-12 on patrol identifying and mapping wildland fires in CesiumJS. Courtesy Lockheed Martin.
JAM brings together ground, air, and satellite data from NIFC, NOAA, and ADS-B flight trackers (as seen in the image above) and combines them in CesiumJS. State agencies might also have their own assets (e.g., thermal imaging from multi-mission aircrafts) that can be added. The application uses Cesium World Terrain and Bing Maps Aerial imagery streamed as 3D Tiles via Cesium ion, to provide incident commanders and staff an accurate 3D map to enhance the common operating picture. CesiumJS was built for interactive web apps and streaming massive amounts of data—it is lightweight and performant, vital for users receiving rapid updates in the field.
Work on the JAM prototype began in December 2023, and efforts on the current demo started in March 2024. Lockheed Martin, which employs 122,000 people worldwide, assigned five developers on this project. The company selected Cesium and its real-time update capabilities as an efficient, effective tool for visualizing historical wildland fire data. Lockheed Martin continues to leverage commercial partnerships to accelerate development of JAM.
Wind and weather information for an active fire area in CesiumJS. Courtesy Lockheed Martin.
Demo of suppression line visualization in CesiumJS. Courtesy Lockheed Martin.
The crew behind JAM is working toward including all active wildland fires in the US. The alpha version was completed in May 2024, and beta is slated for this summer.
You can fuse your own data with high-resolution terrain and imagery with Cesium ion; get started with a community account.