Industry Insights

AI and Machine Learning Sharpen Satellite Imaging

2020-09-30  |  4 min read 

Imaging advances play a key role in defense strategies across the globe. Synthetic aperture radar (SAR) imaging, which is rapidly advancing, covers diverse orbits while providing visibility through previously challenging atmospheric conditions. Yet imaging satellites provide only one data source. Companies providing global monitoring services also use data from sources like Internet of Things (IoT) networks, ground sensors, aircraft, and drones. By adding machine learning and artificial intelligence capabilities, they are providing high velocity insights that were not previously possible.

BlackSky, for example, uses advanced analytics provided by its Spectra AI platform to improve its geospatial intelligence and satellite imagery. Recently, the company completed initial system checkouts and began delivering imagery from its fifth and sixth satellites just 58 hours after they launched. They collected three images in quick succession over Port Elizabeth, South Africa on August 10. BlackSky claims to offer a near-real-time view of the port’s commercial activity. In fact, Spectra AI enables BlackSky to count the container stacks in the yard, identify the container and bulk carrier ships that are present, and count the vehicles awaiting export in the car terminal. It also can track ground vehicle movements and even determine the cargo holds that are currently being loaded.

Monitoring disaster areas and nature also is a key application for such systems. As California faces another series of wildfires, scientists at Salo Sciences are using AI and satellite images to create a highly detailed view of the state’s forests. As summarized by FastCompany, “The tool, called the California Forest Observatory, uses AI and satellite images to create an ultra-detailed view of the state’s forests—aiding work to prevent the type of catastrophic megafires that the state is experiencing now.” Experts trying to combat the state’s wildfire problem need detailed, up-to-date information about the forests so they can better predict how fast and in what direction fires will spread. Such information also gives them a chance to remove hazardous, dense fuels for the fire to burn.

According to the FastCompany article, firefighters can use the tool to predict how current fires may spread as they’re burning. But just as critically, the state can also use the map to plan forest management to prevent future megafires. The Salo Sciences tool uses data from Planet, which controls over 100 satellites that spin in a line around the Earth. They collect images at a resolution and order of magnitude that is much higher than previous generations of satellites.

With machine learning capabilities, results are more quickly achieved. Launched satellites quickly begin imaging, allowing customers to access and use key information. Machine learning and AI then adopt a new role, adding detail and intelligence to images in a way that strives to be predictive. From natural disasters to national security risks, these imaging solutions provide increasingly critical intelligence to help humans form a better educated response to potential dangers.

Keysight supports satellite development over the design lifecycle. Check out our brochure for an overview of our end-to-end solutions. We also cover a variety of industry and design topics in our short “Space Talks.” Catch an episode today.