Causal Machine Learning in Aerospace PhD

This PhD project will investigate the recent field of study of Causal Machine Learning, which aims to modify and augment Machine Learning by using Causal Analysis techniques as a way to solve its limitations. The project will be focused in the aerospace domain with applications ranging from unmanned aerial vehicles, through helicopters, electric vertical take-off and landing aircrafts, robotics and automation, to space and planetary exploration among many others. The research will involve the development of new techniques, comprehensive modelling and software implementation, followed by the construction of experimental prototypes to validate and demonstrate the results. This project is initially directed to self-funded PhD candidates, but possible sources of funding may be available, so we encourage all interested candidates to contact us.

The explosive development of the machine learning field in recent years is limited by a problem intrinsic to its own design. Current machine learning techniques are built to learn how to perform tasks by identifying patterns and correlations by repeatedly observing how to solve those tasks. This implies that these techniques are by design oriented towards imitation rather than reasoning. In other words, they are ineffective in understanding that correlation does not imply causation. This design flaw is clearly exemplified by recent large language models such as ChatGPT that are able to mimic human language surprisingly well, yet fail remarkably at very simple logical reasoning.

In this project, we will investigate the recent field of study of Causal Machine Learning, which aims to modify and augment Machine Learning by using Causal Analysis techniques. These techniques were until now mainly limited to the field of statistics, but in recent years more and more researchers are applying them to Machine Learning as a way to solve its limitations.

The aim of the project is to develop and apply these techniques in the aerospace domain. Being a field of such complexity and precision, this is a perfect scenario to try to expand the potential of machine learning techniques. Within this field, we have different application possibilities, ranging from unmanned aerial vehicles, through helicopters, electric vertical take-off and landing aircrafts, robotics and automation, to space and planetary exploration among many others.

Some of the Causal Machine Learning tasks we focus on include: development of autonomous drone piloting systems capable of identifying the different elements of the environment and how they relate to each other when designing their routes and planning their tasks, systems capable of learning by observing humans performing tasks and identifying the reasoning applied to them, self-driving cars that understand their environment in a robust and rational way and whose actions are explainable, development of dynamic physical models that allow counterfactual reasoning of highly complex systems such as plane engines, automated robotics that allow fully autonomous exploration and modelling of unknown environments or environments where no human intervention is possible, such as nuclear power plants or planetary exploration, etc. The specific focus of the project will be determined jointly with the student participating in the project.

This research project will: 

• Investigate and define the current state of the art for Causal Machine Learning and its applications in the aerospace domain. 
• Define a relevant task based on the previous analysis and develop new methodologies for it. 
• Apply the developed methodologies to simulated environments as well as to real prototypes that will be built and tested in the university installations.

This research thus will involve the development of new techniques, comprehensive modelling and software implementation, followed by the construction of experimental prototypes to validate the results and demonstrate its contributions, as well as to identify any practical issues.

We invite students interested in joining this project, focused on an area of Machine Learning that may be key to bringing about the next revolution in Artificial Intelligence, and to do so with a focus on one of the most complex and interesting technological areas such as the aerospace field. Cranfield University works with over 1,500 organisations, including the leading global aerospace companies such as Airbus, Boeing, Rolls Royce, and Thales, and public agencies such as NASA and ESA, and has one of the largest supercomputing centres in the UK as well as its own airport at the university. This project is based at the Digital Aviation Research and Technology Centre (DARTeC) . This makes it an ideal research centre and environment for the development of this project.

This project is initially directed to self-funded PhD candidates, but possible sources of funding may be available, so we encourage all interested candidates to contact us.

• Application deadline24 Jul 2024
• Award type(s)PhD
• Start date30 Sep 2024
• Duration of award3 years or 6 part time
• EligibilityUK, Rest of world
• Reference numberSATM466

Professor Weisi Guo - is the Director of the Smart Living Grand Challenge and Head of Human Machine Intelligence Group at Cranfield. He is also a Turing Fellow with The Alan Turing Institute. He has been PI on £6.5m and investigator on over £19m of research funding. He has published 130+ journal papers (total IF 710+) and 80+ IEEE/ACM conference papers, with over 5700+ citations (h-index 39). This includes a Nature, Nature communications, Nature Machine Intelligence, Nature Comp.Sci., a top 10% cited paper in PLOS ONE, and several cover issues in Royal Society and IEEE journals. He currently serves as editor on several IEEE & Royal Society journals, and is a Full Member of the EPSRC peer-review college, as well as reviewing for UKRI FLF, ESRC, MRC, Royal Society, and Leverhulme. His research has won several international awards, including IET Innovation in 2015 and Bell Labs Prize Finalist in 2014 and Semi-Finalist in 2016 and 2019.

Dr Miguel Arana-Catania - is a Research and Teaching Fellow in Causal Inference and Verification in Machine Learning. He works at the School of Aerospace, Transport and Manufacturing (SATM) as a member of the Human Machine Intelligence Group. He teaches several modules of the Applied AI MSc and Autonomous Vehicle Dynamics and Control MSc. He obtained his MSc and PhD degrees in Theoretical Physics from the Universidad Autonoma de Madrid (UAM) and the Institute for Theoretical Physics (IFT) UAM-CSIC. He has previously worked in Natural Language Processing at the Alan Turing Institute and at the Department of Computer Science, University of Warwick.

Applicants should have a first or second class UK honours degree or equivalent in a related discipline. This project would suit candidates with a sound background in engineering, physics, computer science, or related disciplines.

This is a self funded PhD.

Research students at Cranfield benefit from being part of a dynamic, focused and professional study environment and all become valued members of the Cranfield Doctoral Network.  This network brings together both research students and staff, providing a platform for our researchers to share ideas and collaborate in a multi-disciplinary environment. It aims to encourage an effective and vibrant research culture, founded upon the diversity of activities and knowledge. A tailored programme of seminars and events, alongside our Doctoral Researchers Core Development programme (transferable skills training), provide those studying a research degree with a wealth of social and networking opportunities.

For further information please contact: 

If you are eligible to apply for this studentship, please complete the online application form.

This vacancy may be filled before the closing date so early application is strongly encouraged.

Please ensure that your fully completed online application form is submitted by the application closing date. All requested documentation should be uploaded to the online form before submission. Note, your application will not be considered unless all relevant documents have been uploaded. For more information please visit Applying for a research degree.