Human-in-the-loop machine learning for drone-assisted Structural Health Monitoring

Project Description:

This PhD scholarship is offered by the EPSRC CDT in Offshore Wind Energy Sustainability and Resilience, a partnership between the Universities of Durham, Hull, Loughborough and Sheffield.  The successful applicant will undertake six-months of training with the rest of the CDT cohort at the University of Hull before continuing their PhD research at Loughborough University. The project is part of a PhD Research Cluster, Reliability and Health Monitoring Cluster.

Inspections of offshore wind turbines, such as identifying damage or ice on turbine blades, anticipating its effects and making decisions on maintenance and repair, as well as estimating remaining useful life (RUL), is an important part of extending the lifetime of a wind turbine as well as the power that can be generated from it. While both tasks are often driven by experts, public data on environmental, meteorological or physical conditions, in combination with satellite and / or climate data, can help make predictions for new, unseen conditions.

The latter is particularly relevant when data is sparse. While public data exists on general environmental conditions and turbine power yield, data around specific combinations of operational and environmental conditions is not always readily available — this is particularly the case for new generations of floating or far-offshore turbines, which are much harder to reach and inspect than previous generations much closer to shore, and for which less historical data is available.

This project aim for two key research advances: first, the development of a new human-in-the-loop active learning framework [2, 3], which uses conversational AI to negotiate key decisions related to turbine inspection and maintenance with a human expert [4, 5]. This can be based on a deep reinforcement learning framework, which interactively optimises key performance indicators in the form of a human-expert informed reward function. Second, we aim for the integration of low-energy machine learning algorithms, so that the resulting AI model can run on a variety of devices, including UAVs (e.g. drones) that may be used in turbine inspection.

The overall aim is the design of a portable learning system that creates a profile of wear and tear of turbines given the environmental, meteorological and physical conditions they operate under. Such data can inform structural health monitoring for offshore wind turbines or help plan new offshore sites, via estimation of power yield in relation to environmental conditions and logistical constraints, such as closeness to shore, shipping routes etc.

Training and skills

You will benefit from a taught programme, giving you a broad understanding of the breadth and depth of current and emerging offshore wind sector needs. This begins with an intensive six-month programme for the new student intake, drawing on the expertise and facilities of all four academic partners. It is supplemented by Continuing Professional Development (CPD), which is embedded throughout your 4-year research scholarship.

Entry requirements

If you have received or expect to achieve before starting your PhD programme a First-class Honours degree, or a 2:1 Honours degree and a Masters, or a Distinction at Master’s level a degree (or the international equivalents) in computer science, engineering, physics or mathematics and statistics, we would like to hear from you.

Recruitment has closed for this project and applications are being assessed for September 2025 entry.

If you have an queries about the research project please contact Prof Nina Dethlefs via n.dethlefs@lboro.ac.uk. You may also address enquiries about the CDT to auracdt@hull.ac.uk.

Watch our short video to hear from Aura CDT students, academics and industry partners:

Funding

The CDT is funded by the EPSRC, allowing us to provide scholarships that cover fees plus a stipend set at the UKRI nationally agreed rates. These have been set by UKRI as £20,780 per annum at 2025/26 rates and will increase in line with the EPSRC guidelines for the subsequent years (subject to progress).

Eligibility

Our funded Doctoral Scholarships are available to UK Students. The advertised CDT scholarships in this current recruitment round are available to Home (UK) Students only as the CDT has reached the annual cap, set by the funding council (UKRI EPSRC), on international student recruitment for the 2025 intake. To be considered a Home student, and therefore eligible for a full award, a student must have no restrictions on how long they can stay in the UK and have been ordinarily resident in the UK for at least 3 years prior to the start of the scholarship (with some further constraint regarding residence for education).

Guaranteed Interview Scheme

The CDT is committed to generating a diverse and inclusive training programme and is looking to attract applicants from all backgrounds. We offer a Guaranteed Interview Scheme for home fee status candidates who identify as Black or Black mixed or Asian or Asian mixed if they meet the programme entry requirements. This positive action is to support recruitment of these under-represented ethnic groups to our programme and is an opt in process.

Find out more

How to apply

Recruitment has closed for this project and applications are being assessed for September 2025 entry.

Interviews will be held online with an interview panel comprising of project supervisory team members from the host university where the project is based.  Where the project involves external supervisors from university partners or industry sponsors then representatives from these partners may form part of the interview panel and your supplementary application form will be shared with them (with the guaranteed interview scheme section removed). Interviews will take place during early and mid-June. 

If you have an queries about the research project please contact Prof Nina Dethlefs via n.dethlefs@lboro.ac.uk. You may also address enquiries about the CDT to auracdt@hull.ac.uk.

References & Further Reading

[1] Chatterjee et al. (2023) Domain-invariant icing detection on wind turbine rotor blades with generative artificial intelligence for deep transfer learning. Environmental Data Science. https://doi.org/10.1017/eds.2023.9

[2] Arulkumaran et al (2017) Deep Reinforcement Learning: A Brief Survey, in IEEE Signal Processing Magazine, vol. 34, no. 6, pp. 26-38, Nov. 2017, 10.1109/MSP.2017.2743240.

[3] Mosqueira-Rey, et al. (2023) Human-in-the-loop machine learning: a state of the art. Artif Intell Rev 56. https://doi.org/10.1007/s10462-022-10246-w

[4] Nishida et al. (2022) Improving Few-Shot Image Classification Using Machine- and User-Generated Natural Language Descriptions. Findings of NAACL, Seattle, USA. 10.18653/v1/2022.findings-naacl.106

[5] Brown et al. (2020). Language Models are Few-Shot Learners. NeurIPS. https://arxiv.org/abs/2005.14165