NUS Smart Sensor Breakthrough: Decodes Fatigue and Stress from Body Signals on the Move

Singapore's fast-paced work culture has led to alarming levels of burnout, with about one in three employees reporting chronic fatigue—one of the highest rates worldwide. This not only hampers productivity but also heightens accident risks in safety-critical roles like driving and aviation. Traditional assessments rely on subjective questionnaires, limiting real-time intervention. A groundbreaking development from the National University of Singapore (NUS) addresses this gap with a wearable smart sensor that decodes fatigue and stress from body signals even during movement.

The NUS innovation promises continuous, objective monitoring, potentially transforming workplace safety and mental health management in Singapore's universities and beyond.

The Urgent Need for Reliable Fatigue Detection

Fatigue and stress contribute significantly to accidents globally. In aviation, fatigue factors into up to 20% of incidents, with surveys showing four in five pilots managing tiredness mid-flight. For road safety, drowsy driving accounts for 13-20% of crashes, amplified in Singapore where long hours plague taxi drivers—studies reveal high fatigue prevalence among them. Locally, the economic toll is steep, underscoring the demand for precise, non-invasive tools.

NUS researchers recognized that existing wearables falter during motion, as muscle twitches and shifts drown out vital signals like electrocardiograph (ECG) readings. Their solution targets this core challenge head-on.

Unveiling NUS's Metahydrogel Artefact-Mitigating Platform (MAP)

The star of this breakthrough is the metahydrogel artefact-mitigating platform (MAP), a soft, skin-like sensor placed on the neck or throat. Unlike rigid devices, this hydrogel conforms perfectly to skin, capturing respiratory patterns—inspiration and expiration shifts—that signal stress or exhaustion.

Development spanned four years: initial sensing tech, then metahydrogel two-and-a-half years ago, followed by a year each on materials and integration. The result? A breathable, stretch-durable patch exceeding human skin's water vapor transmission rate.

Step-by-Step: Dual Noise-Filtering Mechanism

The MAP's ingenuity lies in simultaneous multi-noise suppression:

• Mechanical Vibration Blocking: Self-assembled nanoparticles form periodic bands in the hydrogel, scattering and absorbing vibrations like acoustic panels trap sound—targeting motion-induced noise.
• Ion Speed Control: Biocompatible glycerol-water electrolyte tunes ion mobility, allowing low-frequency heart signals (<30 Hz) to pass while blocking high-frequency muscle noise.
• AI Denoising: A machine-learning algorithm eradicates residual unstructured noise, preserving physiological nuances.

This trifecta yields ECG signal-to-noise ratios (SNR) of 37.36 dB during activity—vastly superior to baselines—and blood pressure deviations under 3 mmHg, meeting ISO 81060-2 standards.

AI-Powered Classification: From Signals to Insights

Fed clean data from multi-day wearables, including simulated driving inducing fatigue, a deep-learning model classifies exhaustion levels at 92% accuracy—double the 64% without MAP. Peak detection in ECG jumps from 52% to 93%, enabling fatigue pattern discernment.

"Our system achieves around 37 dB during daily activities," notes first author Dr. Tian Guo, outperforming smartwatches' 10-20 dB that drop 40% in motion.

Photo by Jorge Ramirez on Unsplash

Research Team and Publication Milestone

Led by Professor Ho Ghim Wei from NUS's Department of Electrical and Computer Engineering, College of Design and Engineering, with Dr. Tian Guo as lead, the work appeared in Nature Sensors on March 24, 2026 (DOI: 10.1038/s44460-026-00055-x). Prof. Ho emphasizes clinician collaboration: "We hope to link data to pathological conditions with clinical accuracy."

This builds on NUS's wearable legacy, positioning Singapore universities as hubs for health tech innovation.

Superiority Over Commercial Wearables

• SNR: 37 dB vs. 10-20 dB (smartwatches).
• Motion Tolerance: No 40% drop; stable contact via conformability.
• Versatility: Suppresses noise in ECG, heart/respiratory sounds, voice, EEG, eye movements.

Reviews confirm wearables excel in controlled settings but struggle ambulatory—NUS MAP bridges this.

Real-World Applications in Safety-Critical Fields

For Singapore's transport sector, where fatigue drives accidents, pilots could receive alerts mid-flight, drivers during shifts. Athletes gain edge via exertion tracking; factories prevent mishaps. In higher education, NUS researchers eye student wellness amid academic stress.

Aviation fatigue stats highlight urgency; this tech could slash risks.

Impacts on Mental Health and Productivity

Beyond safety, continuous tracking enables proactive interventions, curbing Singapore's burnout epidemic. Economic gains: reduced absenteeism, accidents costing billions globally. Universities like NUS foster interdisciplinary solutions, blending engineering and health.

Future Outlook: From Lab to Market

Next: clinician partnerships for validation, industry scaling. Prof. Ho: "Optimise manufacturing for products." Potential in neurophysiological monitoring expands scope.

NUS's iHealthtech drives such transitions, inspiring Singapore's higher ed ecosystem.

Photo by ThisisEngineering on Unsplash

NUS's Role in Singapore's Health Tech Landscape

Singapore universities lead wearables: NTU's AI sensors, NUS's InfinityGlove. Government RIE2025 funds amplify. For academics, opportunities abound in research jobs advancing these frontiers.