Φsat-2: The satellite for Earth Observation with On-Board AI Technology

Φsat-2 mission is to enhance our ability to capture and analyze multispectral satellite imagery (MSI), crucial for precision agriculture, environmental monitoring, and disaster response applications. The mission is a collaborative effort involving a consortium of industry leaders and research institutions dedicated to pushing the boundaries of Earth observation. The consortium led by Open Cosmos includes Ubotica, GGI, CEiiA, GEO-K, KP Labs, and SIMERA.

 

The primary objectives of Φsat-2 include improving data acquisition and processing capabilities directly on the satellite. Using AI techniques, Φsat-2 will optimize data transfer, reduce costs, and significantly shorten the time required to deliver actionable information to Earth. This mission is designed to extract valuable knowledge from raw data, providing rapid insights critical for effective decision-making in various domains. Additionally, Φsat-2 aims to demonstrate the feasibility and benefits of processing data on-board to reduce the raw data that needs to be transmitted back to Earth, thereby accelerating the entire data analysis process and allowing for quicker responses to observed events.

 

KP Labs is instrumental in developing the Cloud Detection Application (App) component of the Φsat-2 mission. This app will preprocess satellite images by identifying and removing cloud-covered areas, ensuring that only precise, valuable data is processed further. By eliminating cloud interference, the Cloud Detection App will enhance the accuracy and efficiency of subsequent AI applications on board, leading to more reliable and timely insights. This feature is an important part of the satellite mission, as clouds can obscure essential features in satellite images, making it challenging to analyze environmental changes accurately. The cloud detection process will help filter out unusable data, optimize the use of satellite resources, and improve the quality of the information gathered.

 

KP Labs’ contribution addresses the challenges of high-dimensional data analysis, transfer, and storage, which are often bottlenecks in deploying multi- and hyperspectral imaging systems. By processing data on board, KP Labs’ technology reduces the need for extensive data transfer to Earth, thereby saving time and resources. This approach optimizes the satellite’s overall efficiency and ensures that critical information is available more quickly for the various applications.

The Φsat-2 mission is expected to deliver its first data set shortly after launch. This data will provide valuable insights into the effectiveness of the on-board AI applications and their impact on Earth observation capabilities. The early data will be critical in fine-tuning the AI models and improving the overall performance of the satellite’s data processing capabilities.

 

For more information about the Φsat-2 mission and KP Labs’ role, visit the ESA website at www.esa.int/Phsat-2 and https://kplabs.space