Duration: 12 months
OHB Hellas, in collaboration with the Fraunhofer-Institute for Integrated Circuits (IIS), proposes a study within ESA’s Open Discovery Ideas Channel to evaluate the applicability of neuromorphic computing—specifically spiking neural networks (SNNs)—for future AI-driven space applications. The study aims to assess the feasibility and benefits of deploying SNN models on Fraunhofer’s SENNA neuromorphic processor, emphasizing ultra-low latency and energy-efficient processing for onboard autonomy. Thanks to its experience from past projects, OHB Hellas will adapt and test algorithms such as object detection, anomaly detection, and image enhancement with SENNA hardware capabilities in mind. A structured approach involving algorithm selection, SNN conversion, simulation-based validation, and performance trade-off analysis will guide the study. A use cases selection phase, early in the project will help to target the most relevant use cases among: anomaly detection in telemetry, active beam steering, and onboard super-resolution and other. The goal will be to demonstrate SENNA’s suitability for such kind of uses focusing on real-time & power constraints.
