Edward Tomanek-Volynets

Email: e.tomanek-volynets.1@research.gla.ac.uk

ORCID iDhttps://orcid.org/0009-0006-2927-5632

Research title: Space Trajectory Design Using Artificial Intelligence

Research Summary

Biography

I graduated with a MEng in Aeronautical Engineering from the University of Glasgow in 2023, with a final year dissertation on optimisation of low-thrust space trajectories. I currently work in the Space and Exploration Technology group of the James Watt School of Engineering, pursuing a PhD in Aerospace Engineering, supervised by Dr. Matteo Ceriotti and Professor Colin McInnes. In addition to my research, I have been involved in delivering a number of undergraduate engineering courses since 2021.

I am a visiting researcher at the European Space Agency's Advanced Concepts Team (ACT), working on mission analysis and design. Specific details of the project which I carry out can be found here.

Current research

My work explores the use of machine learning to quickly optimise spacecraft trajectories in missions with a large number of destinations. This is a combinatorial problem which is extremely (often prohibitively) time-consuming to design using conventional methods, but very useful in contexts such as space debris removal (the necessity of which is becoming quite urgent); so the possibility of training approximation models to learn properties of good solutions is being explored.

My research uses reinforcement learning to tackle this "sequence design" problem; an agent is trained to obtain near-optimal solutions without the need for resource-intensive iterative optimisation. A further problem within multi-target mission design is the calculation of costs of mission legs between individual pairs of orbits (many of which need to be calculated to select the sequence). The use of neural network approximators instead of full trajectory optimisers for these leg costs has become commonplace due to their evaluation speed, but there is still room for improvement on the accuracy, training time and versatility of these tools, so I am also performing research on these techniques.

I am open to collaboration so if you are performing research that links in with my topic, please do not hesitate to send an email.

Broader research interests

Whilst the main focus of my PhD work is on multi-target mission design and machine learning, I have broader interests in a number of areas:

  • High-efficiency space trajectory optimisation/optimal control methods in general, whether conventional or ML-based.
  • Spacecraft control with high specific impulse propulsion, and propellant-less methods like solar sailing.
  • Improving the sustainability of our use of space.

Publications

List by: Type | Date

Grants