EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

ARACHNID: A Neuromorphic Camera Array System for Space Situational Awareness

Sebastián Valdivia, Vicente Westerhout, Daniel Yunge, Esteban Vera

IEEE Transactions on Aerospace and Electronic Systems · 2026 · 62 · 6922-6938

DOI: 10.1109/taes.2026.3669869 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Aerospace and Electronic Systems, 2026. Verfasst von Sebastián Valdivia und 3 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 56 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: FONDECYT REGULAR, ANID ANILLO, ANID BECAS MAGISTER NACIONAL.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The increasing demand for real-time surveillance with large field of view coverage in space situational awareness (SSA) is pushing conventional optical and radar-based solutions in terms of computational, storage, communications and energy capabilities, besides cost. We present the Automatic Real-time Aerospace deteCtion Harnessing NeuromorphIc Devices (ARACHNID) system, a scalable neuromorphic architecture that integrates an event-based camera (EBC) array system with spiking neural networks (SNNs) for real-time detection and tracking of moving space objects. EBCs provide asynchronous detection of brightness changes with microsecond latency, while SNNs enable sparse, event-driven inferences with high energy efficiency. To support training and evaluation of the system, we introduce the event-based vision artificial satellite dataset, a publicly available collection of over 650 labeled event-based sequences of resident space objects (RSOs), recorded under diverse conditions. Field trials demonstrate ARACHNID’s robust detection performance across varying RSOs, achieving an average pipeline latency of 92.5 ms with eight EBCs. Compared to conventional CPU or even GPU-based computing pipelines, ARACHNID delivers superior inference speed and higher energy efficiency in asynchronous operation. By combining neuromorphic sensing and computing, the proposed system offers a scalable low-power framework for real-time, distributed optical space surveillance for SSA.

Förderung

  • FONDECYT REGULAR
  • ANID ANILLO
  • ANID BECAS MAGISTER NACIONAL

Zitieren

APA

Valdivia, S., Westerhout, V., Yunge, D., & Vera, E. (2026). ARACHNID: A Neuromorphic Camera Array System for Space Situational Awareness. IEEE Transactions on Aerospace and Electronic Systems, 62, 6922-6938. https://doi.org/10.1109/taes.2026.3669869

Vancouver

Valdivia S, Westerhout V, Yunge D, Vera E. ARACHNID: A Neuromorphic Camera Array System for Space Situational Awareness. IEEE Transactions on Aerospace and Electronic Systems. 2026;62:6922-6938. doi:10.1109/taes.2026.3669869

BibTeX
@article{Valdivia2026,
  author = {Valdivia, Sebastián and Westerhout, Vicente and Yunge, Daniel and Vera, Esteban},
  title = {ARACHNID: A Neuromorphic Camera Array System for Space Situational Awareness},
  journal = {IEEE Transactions on Aerospace and Electronic Systems},
  year = {2026},
  volume = {62},
  pages = {6922--6938},
  doi = {10.1109/taes.2026.3669869}
}
Kostenlos starten →