Enhancing Space Systems Integrity: a comparison of telemetry-based approaches for satellite PHM

Satellites and space systems play a crucial role in space exploration, especially as interest in the field continues to grow. The substantial costs and investments required for space missions, and the impossibility of directly intervening on the systems in the event of faults or anomalies, necessitate the design and realization of highly reliable platforms. However, the harsh environment they operate in is such that, often, unexpected and undesired events might occur, with potentially catastrophic consequences on the mission objectives. To address this issue, leveraging telemetry data becomes imperative for the early diagnosis of arising anomalies and the prediction of their possible evolution, enabling the implementation of corrective/mitigating actions, albeit difficult, due to the remoteness of the space platforms and the limited range of intervention options. This proactive approach can significantly extend the mission duration, preventing the loss of scientifically or commercially valuable data and functions. Nevertheless, the complexity and multidimensionality of telemetry data, encompassing both sensor measurements and commands, present significant challenges in analysis, underscoring the continued necessity for experts to check system integrity. Currently, most telemetry data processing algorithms rely on predefined thresholds to assess the severity of the anomalies. However, this approach assumes the system already exhibits faults and lacks predictive capabilities to foresee these issues proactively. These limitations pose risks to both operational efficiency and life-cycle management, as accurately predicting the future health state becomes challenging, potentially endangering mission success. Consequently, there is a critical need for advanced automated telemetry data analysis to ensure the success of the missions. To tackle these challenges, this work proposes a framework for the implementation of intelligent prognostics and health management (PHM) algorithms for telemetry data processing, incorporating both machine learning and standard statistical methods. The tools advance a predictive approach aimed at evaluating the health states of space systems by leveraging accurate trend recognition techniques. Furthermore, a comparative analysis between these approaches is conducted to elucidate their respective potentials and limitations comprehensively, especially considering the robustness of these algorithms for space mission reliability.