Trade-Space Exploration With Data Preprocessing and Machine Learning for Satellite Anomalies Reliability Classification

Aug 3, 2026

Satellite reliability is essential for uninterrupted aerospace operations because anomalies can trigger mission failures and substantial economic losses. Many existing anomaly classification approaches struggle with scalability, interpretability, and adaptation across heterogeneous datasets. The study proposes the Trade-Space Exploration Machine Learning (TSE-ML) framework, an end-to-end pipeline that optimizes preprocessing, transformation, normalization, and machine learning. Using the Seradata dataset of 66 years and 4,455 records, the framework evaluates 480 configurations and selects Iterative Imputation, FastText, Robust Scaling, and a Decision Tree, reaching 95.74% testing accuracy. Stratified 5-fold cross-validation and decision-tree interpretability identify key drivers such as age since launch, design life, and orbit category.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages19

Reading Time48 min

AccessFree to download

Abstract

Index Terms

Introduction

Satellite orbit and mission diversity

Space debris and reliability risks

Seradata analysis and motivation

Received 30 January 2025, accepted 11 February 2025, date of publication 19 February 2025, date of current version 28 February 2025. Digital Object Identifier 10.1109/ACCESS.2025.3543813 Trade-Space Exploration With Data Preprocessing and Machine Learning for Satellite Anomalies Reliability Classification ABDUL MUTHOLIB 1,2,(Student Member, IEEE), NADIRAH ABDUL RAHIM1,(Member, IEEE), TEDDY SURYA GUNAWAN1,(Senior Member, IEEE), AND MIRA KARTIWI3,(Member, IEEE) 1Department of Electrical and Computer Engineering, Faculty of Engineering, International Islamic University Malaysia, Kuala Lumpur 53100, Malaysia 2Department of Information System, Faculty of Science and Technology, UIN Syarif Hidayatullah Jakarta, Banten 15412, Indonesia 3Department of Informations Systems, Faculty of Information and Communication Technology, International Islamic University Malaysia, Kuala Lumpur 53100, Malaysia Corresponding author: Nadirah Abdul Rahim ([nadirahabdulrahim@iium.edu.my](nadirahabdulrahim@iium.edu.my)) This work was supported by the Asian Office of Aerospace Research and Development (AOARD) under Grant FA2386-23-1-4073 and Grant SPI23-179-0179 . ABSTRACT Satellite reliability is critical to ensuring uninterrupted operations in aerospace systems, where anomalies can lead to mission failures and significant economic losses. Existing anomaly classification methods often lack scalability, interpretability, and adaptability to diverse datasets. This study introduces the Trade-Space Exploration Machine Learning (TSE-ML) framework, a comprehensive pipeline for satellite anomaly classification that optimizes preprocessing, transformation, normalization, and machine learning stages. Leveraging a Seradata dataset spanning 66 years and 4,455 satellite records, the framework systematically evaluates four data cleaning methods, four data transformation techniques, five normalization strategies, and seven machine learning algorithms across 480 configurations. The optimal configuration, comprising Iterative Imputation, FastText, Robust Scaling, and Decision Tree, achieved the highest testing accuracy of 95.74% with competitive computational efficiency. The Decision Tree model delivered superior accuracy and provided interpretability, revealing critical factors influencing satellite anomalies, such as Age Since Launch, Design Life, and Orbit Category. Stratified 5-fold cross-validation ensured robustness and generalizability of the results. The TSE-ML framework’s transparency and high performance enable actionable insights for improving satellite design, operational planning, and anomaly mitigation. Future research will focus on real-time anomaly detection, integrating satellite telemetry data, and extending the framework to other space applications. This study establishes a robust, interpretable foundation for advancing anomaly classification in aerospace engineering, addressing the dual challenges of reliability and operational efficiency. INDEX TERMS Satellite anomaly detection, satellite reliability classification, trade-space exploration, data preprocessing techniques, machine learning models, seradata dataset, decision support systems. I. INTRODUCTION Satellite systems can be categorized based on their application areas, which include communications, earth observation and remote sensing, navigation, and research. They can also be classified according to their orbital paths: Low Earth Orbit The associate editor coordinating the review of this manuscript and approving it for publication was Mohamed M. A. Moustafa . (LEO), Medium Earth Orbit (MEO), Geosynchronous Earth Orbit (GEO), and High Elliptical Orbit (HEO), as illustrated in Fig. 1. Satellites vary widely in mass, from less than a kilogram to several tons, accommodating a diverse range of missions. Their operational lifespans depend on their mission objectives, with smaller satellites typically lasting a few years, while geostationary communication satellites may

FAQs

What problem does the TSE-ML framework address?

It addresses the challenge of classifying satellite anomalies in a way that improves reliability outcomes while supporting scalability and adaptability to diverse datasets.

How is the TSE-ML pipeline structured?

It provides a comprehensive workflow covering preprocessing, data transformation, normalization, and machine learning stages, then tests many combinations of methods.

Which configuration performed best and what does it provide?

The best configuration combines Iterative Imputation, FastText, Robust Scaling, and a Decision Tree, achieving 95.74% testing accuracy and offering interpretability to highlight factors such as age since launch, design life, and orbit category.

Aug 3, 2026

Upload by Seraphina