Generative LLMs as Personalised Health Coaches: Longitudinal Analysis and Generative Intervention Planning from Multimodal Muscle Fatigue Data

About the Project

These projects are open to students worldwide, but have no funding attached. Therefore, the successful applicant will be expected to fund tuition fees at the relevant level (home or international) and any applicable additional research costs. Please consider this before applying.

This PhD project builds directly upon the successful development of a high-fidelity, multimodal wearable system for real-time muscle fatigue monitoring. This research addresses the critical next step: translating the rich, longitudinal datasets collected into actionable, adaptive, and generative interventions for long-term health management. The existing framework answers, “What is your fatigue state right now?”; this project aims to answer, “Given all my data, what should I do next, and why?”

The core aim is to develop and validate an advanced Large Language Model (LLM) framework that functions as a 'Generative Fatigue Coach'. This AI coach will leverage the large-scale, longitudinal, and multimodal datasets (sEMG, IMU, ECG, heart rate, contextual loads) acquired by the precursor system. The primary focus shifts from real-time detection to longitudinal trend analysis, causal reasoning, and generative intervention planning.

A key novelty will be fine-tuning the LLM to move beyond simple alerts and classifications. The model will be trained to identify long-term patterns, recovery trajectories, and creeping fatigue risks that are invisible in short-term observation. It will be designed for explainable AI (XAI), enabling it to articulate why a specific recommendation is being made by citing specific multimodal data trends (e.g., “Your recovery from high-intensity sessions is 15% slower this week, driven by persistent sEMG median frequency suppression. I recommend replacing tomorrow's planned run with active recovery to avoid overtraining.”).

The generative capability of the LLM will be harnessed to create highly personalised, adaptive plans. Instead of static feedback, the AI coach will generate dynamic rehabilitation programs, adaptive sports training schedules, or optimised work-rest protocols for occupational tasks. This framework will be instantiated within a conversational interface, allowing users to query their long-term progress, discuss subjective feelings of fatigue, and co-create future health plans with the AI.

This research will leverage the established international collaborations (UCL, Imperial, Oxford) to focus on the advanced computational, human-AI interaction, and clinical validation aspects of the system.

Expected Outcomes:

1. A novel data pipeline for structuring longitudinal, multimodal biosignal data for optimal ingestion and fine-tuning of generative LLMs.
2. A fine-tuned LLM framework capable of longitudinal fatigue trend analysis and causal explainability, moving from real-time correlation to deep, personalised insight.
3. A proof-of-concept 'Generative Fatigue Coach' application that produces dynamic, adaptive, and evidence-based intervention plans for users.
4. Validation studies focusing on user trust, adherence, and the long-term efficacy of the AI-generated plans in managing fatigue, improving performance, and mitigating injury risk.

Informal enquiries can be made by contacting Dr R Li (ruizhe.li@abdn.ac.uk).

Decisions will be based on academic merit. The successful applicant should have, or expect to obtain, a UK Honours Degree at 2.1 (or equivalent) in computer science, or a closely related discipline. Candidates should have a strong interest in human–robot interaction, wearable sensing, and LLMs.

Prior experience with Python, Pytorch, and familiarity with LLMs deployment and development will be highly desirable. A good understanding of experimental research methods and data analysis is expected.

We encourage applications from all backgrounds and communities, and are committed to having a diverse, inclusive team.

Application Procedure:

Formal applications can be completed online: https://www.abdn.ac.uk/pgap/login.php.

You should apply for Degree of Doctor of Philosophy in Computing Science to ensure your application is passed to the correct team for processing.

Please clearly note the name of the lead supervisor and project titleon the application form. If you do not include these details, it may not be considered for the project.

Your application must include: A personal statement, an up-to-date copy of your academic CV, and clear copies of your educational certificates and transcripts.

Please note: you do not need to provide a research proposal with this application.

If you require any additional assistance in submitting your application or have any queries about the application process, please don't hesitate to contact us at researchadmissions@abdn.ac.uk

Funding Notes

This is a self-funding project open to students worldwide. Our typical start dates for this programme are February or October.

Fees for this programme can be found here Finance and Funding | Study Here | The University of Aberdeen.