Array Labs’ Post

Recently Adam Smith posted this animation showing a 3D model with the caption, "Data like this from Array Labs and Umbra should be in the hands of every operator in SOCOM for every mission." Wait...data from Umbra? I thought we only did SAR data...so what is going on here? In this specific case, our partners are using non-interferometric techniques to derive 3D data (technically 2.5D data for the experts out there). Sometimes this is referred to as "radargrammetry," an informal play on the much more common photogrammetric techniques now made ubiquitous from consumer drones with cameras. In this example, we capture a site from multiple directions using multiple satellites in quick succession, and by combining complementary angles and high resolution spotlight images, our partner can derive a sub-meter terrain map in a matter of hours. This is especially useful in hard-to-reach places where flying a plane isn't feasible (because of remoteness, hostility, or lack of time). Because we see through clouds and at night and see the whole world every day, we always get the shot no matter the location. A typical site is a few square kilometers and we're testing accuracy of these early returns now with several end-users, mostly against aerial LiDAR ground truth data. So far everyone we've tested with has been pleasantly surprised. We've looked at stockpile measurements, vegetation encroachment on pipelines, and hydrological modeling use cases so far. While this is still a prototype product, this and other more advanced techniques are on our product roadmap. The gold-standard for elevation data generation from space is called "cross-track interferometry," a technique that uses two or more receivers at once to measure millimeter-scale differences in phase data due to precisely known offsets in distance to the target in order to derive 3D structure. For now, we're just using the amplitude data, but soon we'll start using our bistatic formations to test this more exact technique and take full advantage of SAR's many magical properties, like phase information. As always, I'm grateful to work at a place that enables our partners like Array (and dozens more all over the world) to frictionlessly experiment with innovative new ideas and push the whole industry forward along the way.