Artificial intelligence takes control in space: NASA expresses excitement over autonomous satellite operation
Dynamic Targeting: Transforming Earth Observation Satellites
In a groundbreaking development, a new AI-driven technology named Dynamic Targeting is set to revolutionise the way we collect data from space. Developed by Ubotica, Open Cosmos, and NASA's Jet Propulsion Laboratory (JPL), this system is designed to make satellites more intelligent and efficient than ever before.
Until now, satellites have merely acted as passive data collectors, imaging whatever is beneath them and beaming all data back to Earth. However, with Dynamic Targeting, these satellites are transformed into intelligent sensors that can autonomously prioritise and capture significant scientific events.
The key component of Dynamic Targeting is a hyperspectral camera, which allows for detailed analysis of the surface of objects from space. This technology is equipped on HAMMER, an AI-powered satellite launched by Open Cosmos in 2024. HAMMER is a testament to the partnership between the three parties, marking a significant milestone in their collaboration.
Dynamic Targeting works by analysing data onboard the satellite in real time, without the need for human input. The system processes imagery using a small onboard AI processor that evaluates visible and infrared images. This allows the satellite to avoid capturing obscured targets like clouds and instead focus imaging efforts where conditions are optimal.
For instance, during its first flight test aboard the CubeSat CogniSAT-6, the technology successfully detected cloud cover ahead along the satellite’s orbital path and avoided imaging those areas, conserving storage and power for clearer opportunities.
Potential applications of Dynamic Targeting are vast and varied. For example, it can be used for wildfire detection, enabling faster alerts for emergency responders compared to traditional satellite imaging that might take hours or days. Similarly, thermal signatures of active volcanic activity can be autonomously detected and precisely targeted for observation.
Moreover, algorithms are being developed to switch from cloud avoidance to cloud targeting, enabling satellites to hunt for severe storms and deep convective ice storms—rare, short-lived weather phenomena poorly captured by current systems. Using Dynamic Targeting with radar, satellites could track and "stare" at these storm cells as the spacecraft passes overhead, gathering rich data for over six to eight minutes.
Dynamic Targeting also holds promise for atmospheric and climate science, allowing satellites to actively seek out rare atmospheric events instead of relying on chance captures, thereby improving our understanding of Earth's climate.
In essence, Dynamic Targeting transforms Earth observation satellites from passive imagers into intelligent sensors that autonomously prioritise and capture significant scientific events, vastly increasing the efficiency and value of collected data. The technology holds promise for both Earth-focused applications, such as disaster monitoring and climate study, and future missions exploring the solar system autonomously.
The machine learning processor used in Dynamic Targeting and HAMMER was developed by Dublin-based firm Ubotica. In 2021, Ubotica demonstrated real-time AI cloud detection aboard the International Space Station (ISS) as part of a broader research collaboration with NASA's Jet Propulsion Laboratory (JPL).
If the skies are clear, the satellite takes a detailed photo of the surface. However, if clouds obscure the view, the satellite skips the shot to save time, storage, and bandwidth. The satellite can even scan 500km ahead of its orbit and take a preview image. Ubotica's AI analyzes the preview image to check for cloud cover, ensuring that only clear images are captured.
In conclusion, the Dynamic Targeting system, developed by Ubotica, Open Cosmos, and NASA, is set to revolutionise the way we collect data from space. By transforming satellites into intelligent sensors, this technology promises to help scientists get a higher proportion of usable data, opening up a wealth of opportunities for scientific discovery and practical applications.
- The hyperspectral camera technology implemented in Dynamic Targeting, a revolutionary AI-driven system for Earth observation satellites, showcases the intersection of science, space-and-astronomy, and technology, as it facilitates detailed analysis of objects from space.
- Leveraging artificial intelligence, Dynamic Targeting enables these satellites to prioritize significant scientific events autonomously, proving that AI can play a crucial role in space exploration and expanding our knowledge in various scientific disciplines.
