OVERVIEW OF THE PROJECT
The goal of this project was to port our proprietary algorithms for anomaly detection to the onboard FPGA (Field Programmable Gate Array) and run the algorithms on the OPS-SAT telemetry data. OPS-SAT was intended for use as a flying laboratory, carrying various instruments and onboard computing resources, namely the FPGA board. The intention was to enable various experiments in orbit for diverse use cases and to achieve first-flight heritage. This enables easier access into the space industry where the motto was: “Has never flown, will never fly!”. OPS-SAT aims to change that!
Work has been done on porting the needed algorithms to their VHDL equivalents for the onboard Intel Cyclone V and feeding the needed data to the FPGA using ARM. After the processing of the data had been done onboard the satellite, the results were successfully downlinked back to Earth.
The basis for our VHDL code was the algorithms we developed for another client where anomaly detection was of crucial importance. We created multiple IP cores from scratch and connected them to the existing design that was provided for satellite integration. These cores were utilized onboard the ARM chip where a Linux distribution was running. We were required to correctly map the memory of the FPGA by following a template device tree that enabled the Linux system to access the required memory locations for reading from the FPGA and writing to it.

RACE WITH TIME
When we started this endeavor, we knew that the mission was coming to an end since the satellite started deorbiting. However, we were optimistic that we could achieve it in time. Months passed and we were inching closer and closer to completion but we struck some obstacles. As luck would have it, the very last month we faced multiple challenges that impeded us from completing our goal. By spending even more time and energy, seeking support from the OPS-SAT team and our recent visit, we removed the obstacles one at a time. The VHDL project was finished and verified a week and a half (05/08/2024) before the estimated end of the mission. After that, the slow and nail-biting process of hoping the deployment goes well went on. Our very last changes to the code were made just 4 days (05/18/2024) before the end and the very first results of our experiment were downloaded on 05/20/2024 just hours before the mission ended. The satellite disintegrated over Australia on 05/22/2024 around 21:30 UTC, completing its 5-year mission. This redefined our understanding of a photo finish.

LOOKING TO THE FUTURE
We’re grateful for the opportunity that the ESA OPS-SAT team has given us, and honored to be one of the last experimenters on board OPS-SAT. We’d like to thank every OPS-SAT team member who enabled us to conclude the experiment even though the same satellite started deorbiting a few months ago. The OPS-SAT team has spent many hours in the past two months working tirelessly to keep the satellite operational as long as possible while supporting us during the development and deployment of the experiment. We’d like to personally thank Tim Oerther, David Evans, and Vladimir Zelenevskiy who showed patience and willingness to support us as much as they could, even on weekends.
This marks a significant milestone for us as a Croatian company to have an opportunity to run our software in space since the space industry in Croatia is still in its infancy. We hope that this achievement will inspire others to persist in their efforts and see that it is possible to achieve things that may seem impossible as well as to position us at the forefront of the space industry in Croatia.