A research team at UNIST has developed a wearable patch technology that simultaneously analyzes biosignals and environmental information. It is characterized by the application of an on-chip artificial intelligence structure that performs real-time analysis within the device without transmitting data externally.

UNIST announced on the 30th that a research team led by Professors Jae-Jun Kim and Hoon-Eui Jeong has developed a chest-attached patch that detects abnormal conditions by simultaneously sensing biological information such as electrocardiograms and blood pressure, as well as surrounding harmful gases.

This technology is designed using an on-chip AI approach that performs sensor data processing internally within the patch. It is explained that this structure, which transmits only results rather than sending raw data externally as in conventional methods, can reduce communication latency and power consumption.

The patch simultaneously collects biosignals such as electrocardiogram (ECG), photoplethysmography (PPG), bioimpedance (BioZ), and heart sounds (PCG), as well as gas sensor data. Based on this, it can analyze cardiovascular diseases and environmental risk situations together.

The research team applied analog computation-based AI circuits to enable the data processing to be performed internally within the device. The judgment results are transmitted externally via Bluetooth, allowing for remote monitoring of multiple users.

In the performance evaluation, it recorded an accuracy of over 90% in diagnosing hypertension and arrhythmia, and an accuracy of 92.46% in classifying hazardous gas mixtures.

In addition, low-power technology (RPT-PW) that selectively operates optical sensors was applied to improve power efficiency. The research team explained that this technology reduced sensor power consumption by approximately 83%.

The adhesive structure of the patch has also been improved to accommodate prolonged wear. It is designed with a microstructure to enhance skin adhesion while ensuring no residue remains upon removal.

The research team stated that this technology can be utilized for the health management of patients with underlying conditions and the safety monitoring of workers in confined spaces.

This research is scheduled to be published in the Journal of Solid-State Circuits (IEEE JSSC), a journal in the field of circuit design. Anvix Lab, co-founded by the research team, is pursuing the commercialization of the technology.

The research team explained, “On-chip AI-based technology enables stable status determination without an external server,” adding that “it can be utilized in the fields of wearable healthcare and environmental monitoring.”