Research themes

Supervisors

Prof Colin McInnes

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

Asteroids offer to provide material resources to support a range of future space ventures, spanning metals for in-orbit manufacturing and water for in-situ production of propellant. Our prior studies have considered the dynamics of asteroid disassembly using rotational self-energy.

This project will investigate strategies to disassemble rubble pile asteroids using an N-body simulation of the physics of the rubble pile. Disassembly may be required for resource processing, or to reconfigure material for manufacturing structures such as habitats. Such strategies will include free-flying units which remove masses in a serial or parallel fashion, while the rubble pile relaxes into a new minimum energy state after each mass is removed.

 Key research questions include:

• What are the physical limitations on the disassembly of rubble pile asteroids given their gravitational binding energy?
• What strategies can be devised for disassembly using either single or multiple free-flying robotic platforms operating serially or in parallel?
• Can the dynamics of binary asteroids be leveraged to initiate and engineer the flow of material between asteroids?

The project will combine mathematical modelling and simulation to investigate these research questions. Candidates should therefore have a strong aptitude for and interest in mathematical modelling and simulation. The project will be embedded within a large research group pursuing a programme of novel research on emerging space technologies. 

Supervisors

Prof Colin McInnes

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

The in-orbit manufacture of large space structures will underpin a range of new ventures including the fabrication of ultra-large antennae for high bandwidth communications and reflectors for space-based energy services. A key issue will be the development of strategies to actively control both the orbit and attitude of such large structures both during and after fabrication.

This project will investigate novel obit and attitude control strategies for in-orbit manufactured large space structures. Key research questions include:

• How can the orbit and attitude of large space structures be either actively or passively controlled during the fabrication process?
• How does the space environment couple to ultra-large structures, potentially with a time-varying structure geometry?
• What attitude control laws are suitable for in-orbit manufactured large space structures? This will include modelling, simulation and laboratory experiments
• What new strategies can be devised for orbit control of large space structures with strong coupling between orbit and attitude dynamics?

The project will combine mathematical modelling, simulation and some laboratory-scale testing using an air-bearing and Helmholtz cage to investigate these research questions. Candidates should have strong aptitude in mathematical modelling and simulation and an interest in pursuing laboratory experimentation. The project will be embedded within a large research group pursuing a programme of novel research on emerging space technologies.

Supervisors

Prof Colin McInnes

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

Femtospacecraft offer to deliver a broad range of new mission applications spanning space physics, Earth remote sensing and planetary science. A key issue will be the development of strategies to actively control both the orbit and attitude of such small devices.

This project will investigate novel obit and attitude control strategies based on our Mercury 3.5 x 3.5 cm femtospacecraft. The platform comprises a microcontroller with integrated communications, MEMs attitude sensing and 3-axis magnetic actuation. Key research questions include:

• What attitude control laws are suitable for resource-limited femtosatellite? This task will include both modelling, simulation and laboratory experiments
• What is the trade-off between energy/volume used and performance of the attitude control system?
• How can the orbit of a resource-limited femtosatellite be actively controlled and how can the physics of the space environment be leveraged for such tasks?
• How can spatial patterns be formed in swarms of large numbers of such devices to enable new applications of femtosatellite technology?

The project will combine mathematical modelling, simulation and some laboratory-scale testing using an air-bearing and Helmholtz cage to investigate these research questions. Candidates should have strong aptitude in mathematical modelling and simulation and an interest in pursuing laboratory experimentation. The project will be embedded within a large research group pursuing a programme of novel research on emerging space technologies.

Supervisors

Dr Matteo Ceriotti

Dr Kitty Meeks

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

Multi-target space missions (asteroids, debris removal, satellite servicing) are an attractive solution to visit multiple bodies, increasing the scientific return and reducing the cost. Such missions have the potential to be disruptive to lower the cost of both space exploration and increase space sustainability.

The design of a mission to multiple targets from a set is a very challenging problem, combining the spacecraft and trajectory design with the selection of the targets and their order (mixed continuous-discrete optimisation problem). The number of permutations grows exponentially with the number of targets, and it becomes very rapidly infeasible to systematically assess all possible sequences.

This project will leverage recent progress in the design of efficient parameterised algorithms to process temporal networks in order to identify optimal or close-to-optimal sequences (in terms of fuel consumption, mass, time, etc.) without needing to explore the entire search space.

The problem of selecting the best sequence of targets can be formulated as a variant of the Travelling Salesperson Problem, one of the best-known computational problems for which no efficiently scalable algorithm is known in general.  However, there has been substantial work on efficient algorithms to solve this problem either exactly or approximately in special cases; we will draw on recent progress in algorithm design for temporal networks, to develop efficient algorithms to identify (near-)optimal sequences of targets. We will also develop conceptual extensions to the existing mathematical formalism for temporal networks, to better model the continuous changes within the space environment. Later, we will aim to extend our optimisation algorithms to find sets of good solutions rather than just one, allowing an informed choice to be made.

This project will substantially improve the design of multi-body space missions, making use of state-of-the-art techniques in parameterised algorithm design to avoid the need for exhaustive search, thereby reducing the computational time and allowing to use higher-fidelity trajectory and spacecraft models, or exploring more trajectories, or all of these.

This project will also generalise the standard model of temporal networks, since the time needed to traverse an edge can vary continuously over time, opening new theoretical research directions with numerous other potential applications in planning and scheduling applications, not only limited to multi-body space missions.

Supervisors

Dr Matteo Ceriotti

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

Recent research has shown that artificial intelligence can be used to aid the design of space trajectories. Instead of solving a complex and time-consuming optimisation problem, machine learning can be trained to provide an estimate of the cost of a transfer instantaneously. This can be used in preliminary mission design, or in multi-target missions, where the fast computation of a high number of trajectories is necessary. Currently this relies on a database of pre-computed optimal transfers, that can be used to train an artificial neural network.

This PhD project will delve into machine learning and reinforcement learning for space trajectory design, with the main aim to create a framework where an agent would automatically learn about optimality of solutions iteratively, taking away the need to perform any optimisation at all. It will also expand to include the selection of possible mission targets (bodies or orbits), and systems design of the spacecraft itself – very often, the design of the trajectory relies on propulsion system, and vice-versa, the selection of propulsion system relies on the trajectory. Integrated mission-system design will be tackled with machine learning.

Applications include multi-body missions, where thousands to millions of possible trajectories have to be evaluated, such as interplanetary multi-asteroid rendezvouses, and multiple active debris removal missions. However, fast trajectory optimisation is also used in the preliminary phases of the mission design, where the satellite systems (and propulsion system in particular) are not frozen, and shall be selected and optimised together with the trajectory itself.

The ideal candidate will have a background in computer science, artificial intelligence, machine learning, with a strong interest for mathematical modelling and space systems. Alternatively, the candidate can have a background in space systems engineering, with a strong interest (and preferably experience) in artificial intelligence.

References

• Low-thrust, multiple NEA mission design with sample return to Earth using machine learning
• Multi-objective optimization of low-thrust propulsion systems for multi-target missions using ANNs
• Design of multiple space debris removal missions using machine learning

Supervisors

Dr Matteo Ceriotti

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

Quantum computing one of the most important emerging technologies: a step change in our ability to solve difficult problems, in the same way conventional computers have been in the sixties. Conventional computers rely on bits, which can carry on/off information; quantum computers use quantum bits, or “qubits”, which can represent several states at once, exploiting the superposition effect of quantum theory. This allows them to work much faster than conventional computers, and adding more qubits make quantum computers exponentially faster, allowing them to solve problems that are so difficult that are out of reach for ordinary calculators.

In the space mission design, the trajectory design problem is a difficult one, even more so when multiple bodies and/or targets have to be selected from a set (e.g. multiple planetary swing-bys, multiple moon or asteroid tours, multiple satellite servicing and/or disposal): this creates a mixed combinatorial-continuous problem, where the combinatorial part is (broadly speaking) a variant of the classic Travelling Salesperson Problem (TSP), to select the sequence of bodies/targets, and in order to evaluate each sequence, a continuous optimisation sub-problem is to be solved. Quantum computing has the potential to dramatically improve the solution of this problem, my exploiting the superimposition of multiple possible paths at once.

As progress is being made into the hardware to make functional quantum computers, scaling up the number of qubits, this PhD will explore the formulation and solution of space mission design problems through a quantum computing. We aim to answer the following research questions:

• What quantum computing framework(s) can be used for space mission trajectory design?
• How can we leverage on and inject quantum computing to the space mission trajectory design problem, particularly when multiple bodies/targets are involved?
• How can trajectory design problems be encoded through a quantum algorithm?
• To what extent a full trajectory design problem can be implemented as (and take benefit from) a quantum algorithm?

Ultimately, we will assess to what extent, injecting quantum computing into the optimisation problem, we obtain a quantum advantage, both in terms of optimality of solution, and computational cost, for this specific application (narrow advantage).

The ideal candidate will have a background in computing science or similar discipline, with a strong interest in space technology and exploration, or vice-versa a background in space trajectory design with strong interest in computing science and programming.

Supervisors

Dr Kevin Worrall

Dr Gerardo Aragon Camarasa (School of Computing Science)

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

With the current push towards space for both private and government organizations, and the recent increase on initiatives to the industrialization of space, there will be an important need for humans to be supported by robotic systems. Understanding and mastering the unique properties that will intervene in the robot behaviour is essential to offer a fully autonomous robotic system which will be expected to work with no human intervention while being robust, accurate and responsive.

The work will consider the different advantages of both traditional and AI-based control methodologies to support the development of a vision-based control system that is able to control a robot manipulator within the space environment during in-orbit assembly tasks. The expected outcome of this work will be a simulation environment of a suitable setup and a practical real-life implementation.

This project will engage with recent research studies on the field on autonomous robotics, building in-orbit structures, satellite assembly and support studies on manufacturing in space. This project can also engage with users beyond space, with advanced manufacturing research being a potential area to explore.

Background in either control engineering mechatronics, computing science, and/or space engineering is highly recommended. In order to be eligible to apply for the School of Engineering Scholarship, an excellent CV is required.

RESEARCH LINES

This project explores the following lines of research:

• Robotic arms for manufacturing in space
  
  This line of research focuses on the analysis of the dynamics, kinematics, and grasping methodologies of the robotic arms while on orbit. This addresses problems related to autonomous robotics, target capture strategy, tackling a moving orbiting object, mathematical approach to the robotic arm dynamics, and contact forces. In addition, the major physical interactions while executing tasks on orbit such as building in-orbit structures, satellite assembly, and space manufacturing, will be considered.
   
• Approaches for controlling robotic manipulators in space
  
  This line of research focuses on the analysis and exploration of traditional and AI-based control methodologies, intelligent control algorithms, and an integrated vision-based control system. This addresses problems related to the vision system embedded in the robot, environment simulation, and parameters such as speed, torque, vibration, and attitude disturbance.

Supervisors

Dr Matteo Ceriotti

Dr Kevin Worrall

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

Aerial imaging (using drones or satellites) is increasingly becoming essential: pictures from above are today used for mapping, security, disaster prevention and monitoring, weather forecast, traffic monitoring and routing, intelligence, environmental monitoring and planning, and many others. Spatial and temporal resolution are the two main metrics of an imaging system, and very often one is to be traded off for the other: aircraft and drone photography can often provide higher spatial resolution than satellites, but at the cost of a lower temporal resolution or limited coverage area and time window. Conversely, satellites can repass the same large area of interest frequently, and for many years, but due to their altitude, the spatial resolution is lower.

Depending on the purpose of imaging, one requirement might be more important than the other. The spatial resolution of an image is limited ultimately by the distance of the subject, the optics and the aperture of the lens used, due to diffraction. If the distance of the subject is not negotiable, ground-based systems overcome this limitation using larger-aperture optics (e.g. telescopes) as mass and volume are often not an issue. This is entirely different for Earth-imaging airborne and space-borne systems, where volume and mass are heavily constrained, and in fact are often to be minimised. This limitation has led to using large optics supported by large aircraft and spacecraft. However, disadvantages of this approach are high costs and low temporal resolution.

This research will investigate using a constellation of small satellites, effectively acting as a distributed imaging system. Instead of relying on a single device imaging the nadir, multiple devices (agents) in different locations can cover the same area at the same time. Computer-based image-processing techniques, including the use of neural networks and deep learning, will be used to fuse the source images and generate a super-resolved picture of the area of interest, of a better quality overall than each one of the sources. The ultimate aim of the project is to prove that this is a viable alternative to current distributed Earth imaging.

Background in computing science and/or space engineering is highly recommended. In order to be eligible to apply for the School of Engineering Scholarship, an excellent CV is required.

Supervisors

Dr Matteo Ceriotti

Dr Kevin Worrall

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

Intersatellite links (ISLs) are telecommunication routes between different satellites which allow a swarm or constellation of satellites (or agents) to effectively become a network of relay nodes. ISLs can be used to share data amongst different nodes of a network; one possible aim is to maximise the bandwidth between two specific agents in the network, or between an agent and an external entity (e.g. a ground station in the space scenario). With these links in place, satellites in large-enough constellations can communicate with relevant ground stations in quasi-real-time, regardless of whether the ground station is in line-of-sight and/or range. It is clear that the extent of the usefulness of ISLs depends on the effectiveness of the routing strategy employed. The main difficulty in utilizing ISLs is the fact that in most satellite constellations, the network topology is time-varying; links will constantly be found/lost as each satellite progresses along its own orbit, hence the effectiveness of the routing strategy becomes key to exploiting the availability of ISLs.

This PhD will investigate distributed algorithms for the autonomous optimisation of ISLs within a satellite constellation. Previous research [http://eprints.gla.ac.uk/159120] has looked into the use of Ant Colony Optimisation, a bio-inspired technique that mimics the behaviour of ants foraging for food; the PhD will expand this research and assess and compare the use of other optimisation methods. It will also investigate the effect of constraints introduced into the network (such as unavailability of one or more nodes) and develop techniques to cope with them optimally. One of the paramount aspects to consider is that the system should be able to self-optimise itself (fully-distributed) without the need of a central controlling node. In this way, the loss of one or more agents does not prevent the swarm to continue to find optimal solutions.

The techniques developed for the satellite scenario can readily be extended to other applications with different agents, such as autonomous vehicles, drones, sensors, etc.

Background in computing science, applied mathematics and/or space engineering is highly recommended. In order to be eligible to apply for the School of Engineering Scholarship, an excellent CV is required.

 

Supervisors

Prof Colin McInnes
Dr James Beeley
Dr Kevin Worrall

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

Early work on biomorphic autonomous spacecraft considered the use analogue circuits to mimic simple spiking neural networks. It has been shown that such biomorphic systems can demonstrate quite complex emergent behaviour and can be robust to failures. While our work on 3x3 cm PCB-satellites currently uses conventional microcontrollers, the use of biomorphic control may enable even smaller, yet capable devices.

This project will firstly investigate the use of biomorphic control for ultra-small, centimetre-scale micro-spacecraft and then further develop our ideas to consider a large networked swam of devices. Key research questions include:

• How can low-level behaviours be embedded in individual centimetre-scale micro-spacecraft; for example de-tumbling, Sun-pointing, target-pointing and orbit control?
• How can interaction between the low-level biomorphic control of members of a large swarm of such devices lead to emergent, complex high-level behaviour?
• What niche applications can be foreseen which leverage the benefits of biomorphic control while competing against the performance of conventional spacecraft swarms?

The project will combine modelling, simulation and laboratory-scale testing to investigate these research questions. Candidates should have an interest in modelling and simulation and an enthusiasm for laboratory experimentation. The project will be embedded within a large group pursuing a programme of novel research on emerging space technologies.

 

Supervisors

Dr Matteo CeriottiProf Colin McInnes

Funding

Currently unfunded. Please consult the Postgraduate Research section for information on applying for support. 

Description

The design of a space mission starts with the design of a spacecraft trajectory that allows for all the mission goals to be achieved. Traditionally, spacecraft have a propulsion system, which provides the impulse necessary for transfer and orbit control. The technologies mostly used today are high-thrust chemical rockets and low-thrust solar electric thrusters, however other intriguing concepts, such as solar and electric sailing, are being investigated and demonstrated. The types of space trajectories depend heavily on the propulsion technology used, and so do the design methodologies, to a certain extent. Hybrid propulsion is the use of two different propulsion technologies on the same spacecraft, and it can include, for example: low and high thrust, low thrust and sailing. At the increased cost of carrying two propulsion systems, the two of them can be used cooperatively in order to generate novel missions and trajectories that are unfeasible otherwise, or provide cheaper options for existing missions. The design of hybrid-propulsion trajectories will also require the use of new dynamical models and optimisation techniques in order to be able to fully explore their potential.

This project will investigate both the development of new numerical tools to design and optimise hybrid trajectories, and the consequent application to future space missions. Possible missions of interest will include debris mitigation, asteroid orbit manipulation, and more generally, interplanetary transfers, particularly in multi-body environment, where the benefit of different propulsion systems can be exploited most. The PhD will involve both analytical and computer-based (numerical) research, preferably in the MATLAB environment.

The ideal PhD candidate will have a degree in aerospace engineering or applied mathematics, and an excellent track record, preferably including evidence of outstanding research, such as previous awards and/or publications.