RoBo6: Standardized MMT Light Curve Dataset for Rocket Body Classification
Daniel Kyselica, Marek Šuppa, Jiří Šilha, Roman Ďurikovič
TL;DR
RoBo6 addresses the lack of standardized benchmarks for rocket-body classification from light curves by introducing a six-class preprocessing-standardized dataset derived from the MMT database. The paper defines gap-based splitting, quality filtering, and a macro F1 evaluation protocol, and benchmarks five models, finding Astroconformer to perform best. The dataset on HuggingFace offers a consistent platform for fair comparisons and rapid advances in space-object characterization. The work contributes a practical benchmark to advance safe and sustainable space operations.
Abstract
Space debris presents a critical challenge for the sustainability of future space missions, emphasizing the need for robust and standardized identification methods. However, a comprehensive benchmark for rocket body classification remains absent. This paper addresses this gap by introducing the RoBo6 dataset for rocket body classification based on light curves. The dataset, derived from the Mini Mega Tortora database, includes light curves for six rocket body classes: CZ-3B, Atlas 5 Centaur, Falcon 9, H-2A, Ariane 5, and Delta 4. With 5,676 training and 1,404 test samples, it addresses data inconsistencies using resampling, normalization, and filtering techniques. Several machine learning models were evaluated, including CNN and transformer-based approaches, with Astroconformer reporting the best performance. The dataset establishes a common benchmark for future comparisons and advancements in rocket body classification tasks.