Revolutionising Digital Twin and Machine Learning Approach for Space Tracking Sensor Network Management

Description of field of research:

The increasing volume of satellites and debris orbiting the Earth has made the space environment congested, contested, and competitive, necessitating effective tracking and controlling capabilities. However, lack of structured tracking and data processing approaches may limit the tracking and controlling capabilities on the mounting volume of satellites and debris orbiting Earth. A network of sensors, e.g., electro-optical telescopes, is used to track the large population of resident space objects (RSOs), formulated as a multi-objective optimisation problem (Cai 2020). The optimisation approach used for load on sensor network must address the evolving nature of RSOs in the SSA context.

The ongoing fourth industrial revolution, known as Industry 4.0, has sparked a focus on data utilisation for system performance monitoring and optimisation. Equipped with data analytics approaches such as machine learning, the digital twin can serves as a digital replica of the RSO system that dynamically interacts with the real-world tracking system and offers services such as system optimisation, simulation, and scenario analysis.

This project aims to customise the generic approach for developing a systems-of-systems (SoS) architecture (Abdoli 2022, Abdoli 2019) and develop a novel digital twin-based approach to address RSO motion and sensor allocation complexity for search and catalogue maintenance of multiple RSOs.