Simera Sense, a leading European provider of high-performance optical payloads for space-based imaging, today announced that it is developing a combined edge and cloud AI capability with the aim of providing rapid intelligence at a global scale. In fields such as disaster management, climate monitoring and defence, events can unfold quickly across vast areas. Responding quickly can be vital, and delays can prove extremely costly.
Today, satellites capture huge amounts of data, but sending that data to Earth and turning it into useful information can take days. Simera Sense is making this process faster and easier, cutting the delay between capturing an image and putting it to use to just minutes or seconds.
This will be particularly useful for very-high-resolution (VHR) Earth observation, where satellites capture details smaller than one metre in size. As satellite images become more detailed and numerous, AI can help identify important features, objects, activities, and changes more quickly, allowing customers to act on the information rather than simply collect high-resolution images.
Simera Sense is developing a distributed operating system that links its own hardware in space with cloud services on the ground. This will give customers rapid access to satellite imagery and information drawn from it using AI.
On orbit, Simera Sense's operating system and AI models will run on radiation-tolerant computing hardware based on AMD's second-generation Versal technology. The edge computer is being developed tightly and optimally integrated with Simera Sense's new sub-1-metre-class imagers so that the imaging and on-board processing systems are closely integrated. This approach is designed to support the processing demands of next-generation high-resolution imaging, enabling relevant information to be identified closer to the point of capture and potentially reducing the volume of data that has to be sent to the ground. The new imagers are expected to launch in 2028.
Simera Sense's distributed system will also run in hosted environments on the ground, giving data analysts the tools to train, test and deploy models and algorithms that are agnostic to where processing takes place – in space or on the ground. By combining onboard processing with cloud-based computing, the system will support tasks ranging from the rapid detection of specific objects or events to the more detailed analysis of high-resolution imagery. Customers will be able to benefit from AI-assisted insights while retaining the flexibility to apply processing wherever it is most appropriate for their mission. The architecture will also customers to run their own AI algorithms, following Simera Sense’s approach to make space imaging effortless.
AI could help customers use VHR satellite imagery to monitor infrastructure, track changes at important locations autonomously, assess damage after emergencies, and detect unusual activity. By identifying useful information from sub-1-metre imagery more quickly, Simera Sense aims to help customers respond faster and get much greater value from their Earth observation data.
Johann du Toit, CEO of Simera Sense, said: “Our imaging systems are already at the forefront of affordable, high-resolution, space-based Earth Observation. Tightly and optimally integrating on-board and ground-based distributed AI capabilities to assist decisionmakers is the obvious next step.
“As we develop our next generation of VHR imagers, integrating AI processing into the imaging ecosystem will be increasingly important—not only to capture more detailed imagery, but also to help customers extract actionable intelligence from it more quickly.
“We believe these integrated offerings will be a huge step forward towards enabling rapid action from Earth Observation data.”