Textile-integrated electrochemical wearables with edge AI for on-body heat-strain monitoring

Research Theme

Sensor Signal Processing

Autonomous Sensing Platforms

Aim

Build a smart garment sensor that spots heat strain on the body in real time, using very low power (about one-thousandth of a watt) and reaching high accuracy (around 90%) in realistic military settings.

Objectives

1. Build a garment-mounted panel measuring sweat rate, electrolytes (sodium/chloride or bulk conductivity), sweat lactate, and skin temperature.
2. Implement edge artificial intelligence (Tiny Machine Learning) to deliver continuous on-body inference at ≤1 mW average power.
3. Validate in controlled heat/exertion trials, demonstrating ≥90% event-detection F1 and robust operation under sweat and movement.
4. Quantify inter-subject variability (sex, BMI, fitness/acclimation) and perform a 5–10 min per-user calibration (offset/scale or few-shot adaptation); report within- and cross-subject performance.
5. Exposure add-on: Include a chemical-exposure co-monitor (oxidants/irritants) with event qualification (duration–intensity product; recovery slope), targeting F1 ≥0.85 at ≤1 mW incremental power.

Description

We will develop a textile-integrated electrochemical wearable for real-time heat-strain detection at ≤1mW. It measures sweat rate, electrolytes (Na/Cl or conductivity), sweat lactate, and skin temperature, using on-garment edge AI. H₂O₂ serves as an oxidative-stress co-signal and enzyme-assay reporter, not a primary marker. The sensing stack uses printable carbon electrodes with solid-contact ion-selective and enzymatic sensors in breathable laminates; electronics are snap-in and reusable, with peel-and-stick microfluidics as the consumable. An oxidant/irritant co-monitor adds event qualification (duration–intensity product, recovery slope) to distinguish brief surges from prolonged low-level exposure. We will quantify inter-subject variability and conduct a per-user calibration; performance will be reported for within- and cross-subject evaluations. The project aligns with SPADS themes—Sensor Signal Processing and Autonomous Sensing Platforms— and draws on supervisor expertise: Dr E (electrochemical sensing; wearable integration) and Dr Escudero Rodríguez (edge artificial intelligence; signal processing). Output: a prototype targeting ≥90% F1-score in defence-relevant conditions.

Minimum entry qualification - an Honours degree at 2:1 or above (or International equivalent) in a relevant science or engineering discipline, possibly supported by an MSc Degree. Further information on English language requirements for EU/Overseas applicants.

Full funding is available for this position.