top of page

Research

We use artificial intelligence to tackle the challenges of space sustainability. Our work focuses on autonomous satellite operations, predicting how space objects behave, and ensuring the long-term health of the space environment. By advancing AI technologies, we aim to make space safer and more accessible for everyone.

Assembling Spacecraft

Autonomous Operations

How do we keep satellites from colliding in an increasingly crowded space?

We develop advanced autonomous systems to enhance space safety. Our focus is on collision avoidance technology. We are expanding the CASSANDRA framework, a tool for space situational awareness and debris management. By integrating machine learning and Dempster-Shafer theory, CASSANDRA predicts potential collisions with high precision and provides decision support for space operators. It handles uncertainties in space data, ensuring we manage collision risks with confidence.

 

Our current research aims for fully autonomous execution of collision avoidance manoeuvres (CAMs). We use advanced AI and deep reinforcement learning (DRL). This technology calculates optimal and safe CAMs, removing risky decisions before they happen. Our goal is to create a robust, autonomous space traffic management system. We want to manage risky conjunctions effectively and reduce the likelihood of collisions in space.

Spaceship

Space Object Behavioural Analysis

Can we predict the behaviour of every object in space?

Understanding how satellites and debris act is crucial. We combine machine learning with advanced observational data—like light curves and radar imaging—to classify, track, and predict space objects. This helps us prevent collisions and improve space traffic management. Maintaining safety and sustainability in space activities is essential.

Our research uses real observational data and highly realistic simulations. Our AI models reconstruct the motion of space objects and identify potential anomalies. We are developing Foundation Models that fuse data from multiple sensors. These models predict future behaviours and detect unusual patterns.

By understanding space object behaviours, we can take action to prevent collisions. This protects satellites and ensures reliable services on Earth.

Image by SpaceX

Connect with Our Experts

Our innovative research is driven by a team of dedicated experts. Their knowledge and passion fuel advancements in AI for space sustainability. To learn more about the people behind our work, visit our experts' profiles.

Experts Panel

Long-term Space Sustainability

Space is getting crowded. Thousands of satellites orbit Earth, and more launch each year. How can we keep space sustainable for future generations?

We develop advanced tools to assess the environmental impact of space missions. We collaborate with MIT and the University of Strathclyde to create tools like MOCAT (MIT Orbital Capacity Assessment Tool) and NESSY (Network Model for Space Sustainability). These tools help us predict how Low Earth Orbit (LEO) will evolve under different scenarios, considering factors like debris, collisions, and launch traffic.

Using real data and high-fidelity simulations, we forecast space conditions and guide decision-makers in designing sustainable missions. We are developing machine learning models like MOCAT-ML to enhance prediction efficiency. This makes it easier to identify risks and manage space traffic. NESSY allows us to track interactions between space objects and forecast debris growth and collision risks over the long term.

We integrate these tools with our Life Cycle Sustainability assessment platform to evaluate new missions for operational success and environmental impact. By combining innovative AI models, simulations, and risk metrics, we pioneer a new approach to sustainable space mission design. Our aim is to benefit the global space community.

By optimising orbital paths and avoiding congestion, we ensure safer long-term operations. What if we could predict and prevent future space debris growth? Our tools strive to make that possible.

Radio Telescopes
bottom of page