Two computer engineering students at the University of Pittsburgh, Diccia and Nishal Harel, entered history, becoming the first undergraduate students, which deployed their own applications on the satellite, which is in geostationary orbit at high altitude 35 786 km above the ground.
The project was implemented as part of a partnership between the NSF SHREC Center for High-Performance and Sustainable Space Computing and Lockheed Martin, яка надала платформу для експериментів – супутник LINUSS (In-space Upgrade Satellite System). Цей CubeSat, запущений у листопаді 2022 year, is used as an orbital laboratory for testing new software solutions in the difficult conditions of geostationary space, where radiation load and memory limitations create additional challenges.
As part of the collaboration, the students developed two applications based on machine learning. One of them uses algorithms to efficiently compress satellite images, reducing the volume of transmitted data, and the other - autonomously classifies pictures, щоб фільтрувати неінформативні кадри, як-от зображення океанів або хмар. Обидві програми були протестовані на наземній моделі обчислювальної плати ZCU102, after which they were transferred to Lockheed Martin for uploading to the LINUSS satellite.
The launch and successful functioning of applications became an important milestone not only for students, and for the university itself, since this is the first case, when undergraduates from Pittsburgh deployed their own programs on an orbiter in geostationary space. The initiative demonstrates the effectiveness of partnerships between academic institutions and industry in training a new generation of specialists in the field of space technologies and high-performance computing systems.
Source: https://phys.org