AI-Based Satellite Anomaly Detection Using Fused Multi-View Remote Sensing Data: Technologies and Prospects
Year: 2025
Type: article
Abstract: In recent years, satellites have been widely employed in diverse domains such as Earth observation, weather forecasting, and space exploration, generating large-scale data through remote sensing and telemetry.However, these data have primarily been used to analyze surface changes or atmospheric phenomena, while their application to anomaly detection of satellite systems themselves has remained limited.Traditionally, anomaly detection has relied on onboard sensors and ground control centers through Failure Detection, Isolation, and Recovery (FDIR) techniques.Recently, however, increasing attention has been given to detecting anomalous states through cross-observation among multiple cooperative satellites and between satellites and ground stations.In particular, for mission-dedicated satellite constellations such as space-based solar power systems, the necessity of artificial intelligence (AI)-based analysis leveraging remote sensing and telemetry data has been emphasized.This review comprehensively examines recent research trends and representative case studies on AI-based satellite anomaly detection using remote sensing data and provides perspectives on future directions.
Source: 대한원격탐사학회지
Authors: Hoonhee Lee, Jae-Won Kim
Institution: Hanseo University
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
Awards: G3716376693, G7059719119
OpenAlex ID: W4415986635