The objective of this document is the presentation of the final synthesis of the activity “HIPNOS”, Development of a representative HW/SW solution for a high-performance processing platform for Active Debris Removal missions. Implement COTS-based solution as Demonstration of the activity.
HIPNOS includes trade-off study of processing technologies, processing space-grade and COTS devices, processing architectures, trade-off for computer vision algorithms, and demonstrator of the selected solution.
The scope and objectives of the activity are:
- -study/define the high-level architecture of a high-performance computing system for space avionics for GNC in ADR missions
- design one high-accuracy & high-complexity chain of Computer Vision algorithms to support the ADR scenario of e.Deorbit
- select the most appropriate acceleration platform in terms of speed, power, rad-hardness, mass/size, flexibility, future trends
- accelerate the CV algorithms on FPGA or GPU, or DSP, or multi-core CPU to achieve the high- speed processing required in ADR
- develop and demonstrate a preliminary, proof-of-concept system (by using COTS components and high-definition videos) with a representative ADR use-case.
- present the feasibility of implementing demanding algorithms with real-time performance on future space-grade platforms
This work has been performed under the ESA contract No. 4000117700/16/NL/LF. The consortium participating in this project was formed by GMV Aerospace and Defense as prime, and the Greek universities of NTUA and FORTH as subcontractors.