Self-supervised Machine Learning Based Approach to Orbit Modelling Applied to Space Traffic Management

TL;DR

This paper introduces ORBERT, a self-supervised machine learning model inspired by NLP techniques, to improve space traffic management by leveraging large unlabelled orbit datasets for tasks like conjunction screening.

Contribution

The paper presents a novel self-supervised orbit modeling approach, ORBERT, which enhances space traffic management tasks by utilizing unlabelled data for better representations.

Findings

Improved conjunction screening performance using ORBERT.

Self-supervised learning benefits when labeled data is scarce.

Demonstrated effectiveness on orbit time series classification.

Abstract

This paper presents a novel methodology for improving the performance of machine learning based space traffic management tasks through the use of a pre-trained orbit model. Taking inspiration from BERT-like self-supervised language models in the field of natural language processing, we introduce ORBERT, and demonstrate the ability of such a model to leverage large quantities of readily available orbit data to learn meaningful representations that can be used to aid in downstream tasks. As a proof of concept of this approach we consider the task of all vs. all conjunction screening, phrased here as a machine learning time series classification task. We show that leveraging unlabelled orbit data leads to improved performance, and that the proposed approach can be particularly beneficial for tasks where the availability of labelled data is limited.