WP2.C Onboard SW platform for New Space Applications

This task aims at developing innovative SW architectures, with SW relevant technologies, that go beyond the current concept of platform and payload disjoint processing, converging into a single SW platform for New Space Applications.

The SW platform sought is based on a hybrid and parallel computing concept, embedding SW Files System concentrators and memories paradigms such those used into ground data centers. For the processing there will be scalar, vectorial IP's and accelerators derived from programmable logics.

This computing grid based on HW COTS, will be orchestrated with systems based on open sources standards (such as Open MP - Open CL) to maximize efficiency in the allocation of computation for IP’s resources and memories.

The idea is to develop an In-orbit framework with DevSecOps technologies that allows the development, testing and deployment of FLIGHT SW even when the SW platform is in orbit.

The framework engines will be designed to host accelerators for artificial intelligence, capable a very high level of autonomy on board, integrating the most modern AI libraries developed on the new space market (as a reference Tensor Flow, Klepsydra, etc).

This technology will allow AI maneuvers, AI reconfigurations and on-board diagnostics-based AI, with a virtual AI operator that will limit the management costs of complex software from the ground stations.