Sidus Space, Inc., a Space-as-a-Service satellite company focused on commercial satellite design, manufacture, launch, and space-based data collection, announces integration of edge AI capabilities into its planned satellite constellation.
Under an expanded agreement with Exo-Space, Sidus will leverage the FeatherEdge AI platform to provide near real-time intelligence derived from Earth Observation data. This includes opportunities to identify and intervene against environmental emergencies. It can detect unregistered vessels contributing to illegal fishing, track natural disasters such as floods or wildfires in real time, and provide early warnings for events including oil spills or methane leaks.
“FeatherEdge executes novel machine vision algorithms to detect and classify objects and structures within the field of view of on-board imaging systems. FeatherEdge processes raw data directly from sensors and sends down only the relevant information, reducing downlink costs and improving response times to observable events,” said Jeremy Allam, Co-founder and CEO of Exo-Space.
The combination of a rapid revisit rate satellite constellation and onboard high-performance edge AI computing is expected to enable Sidus to offer industry-leading data transmission speed. This combined solution significantly accelerates the speed in which end users can receive analytics after image capture. The integration of edge AI capabilities also allows ground operators to send updated tasking commands, mission objectives (such as regions of interest), and AI models (corresponding to emerging events), enabling the delivery of intelligence that can be adapted to evolving challenges.
“By integrating edge AI into our Sidus fleet of satellites, we expect to be able to provide government and commercial customers with near-real-time intelligence. Edge AI implements advanced analytic methods such as machine learning and artificial intelligence in an edge computing environment. Processing will occur in space, prior to transmitting data back to Earth, and will enable our constellation to learn to prioritize the retrieval of useful intelligence so we can deliver efficient, actionable insights for critical data-driven decisions,” said Carol Craig, Founder and CEO of Sidus Space.