A gas sensor that is both precise and operates at ultra-low power has been developed by combining nanotechnology and artificial intelligence (AI) deep learning, and is expected to be used in various fields such as air quality monitoring, health diagnosis, food safety, and environmental protection in the future.

UNIST (President Jong-Rae Park) announced on the 13th that a research team led by Professor Shin Heung-Joo of the Department of Mechanical Engineering and Professor Kim Jae-Jun of the Department of Electrical and Electronic Engineering successfully implemented an electronic nose that can precisely measure the type and concentration of gas by combining nano-processing and deep learning technology.

The newly developed electronic nose uses a nano-sized heater-based semiconductor gas sensor. While conventional sensors consume a lot of power due to high operating temperatures, this sensor can operate with less than 200 microwatts of power. It is suitable for mobile and IoT devices, and the productivity is also high because the sensor is manufactured based on a semiconductor process.

The high power consumption of existing electronic noses has been reduced through sensor miniaturization. Power consumption has been further reduced by 90% through the introduction of duty cycling technology. Duty cycling is a technique that minimizes energy usage by periodically supplying and cutting power.

The nano-sized heater can heat up to 250 degrees Celsius and cool down to room temperature in just 1/100,000th of a second, enabling smooth gas measurements even under short duty cycles.

The research team improved the existing electronic nose structure that required multiple sensors to operate with a single sensor. The desorption of adsorbed gas on the semiconductor surface is slow compared to the operating speed of the nano heater. Therefore, in the case of rapidly repeated duty cycling, the gas reaction continues even during the short time when the heater is cooled down. Therefore, different signals can be collected during the heater operation and shutdown. These dual signals were analyzed in real time using a convolutional neural network (CNN) to accurately identify the types and concentrations of various gases.

Professor Shin Heung-joo said, “We have been able to solve the limitations of existing electronic noses with a single sensor,” and “It can be easily applied to mobile and IoT devices where miniaturization is essential.”

Professor Kim Jae-jun explained, “We can create an ultra-small gas measurement device that operates on low power, so it can be used in various application fields such as real-time wireless monitoring systems.”

The research results were published in the nanotechnology journal Small and the international journal ACS Sensors. The nanoheater-based gas sensor was published in September 2022, and the electronic nose technology was announced in June 2024, and each was selected as the cover paper. The research was supported by the National Research Foundation of Korea (NRF), the Ministry of Trade, Industry and Energy, and the Ministry of Science and ICT.