"AI Reads Scents"... DGIST Presents Blueprint for Next-Generation 'Electronic Nose' Modeled After Human Olfaction [Reading Science]

A Roadmap for Artificial Olfactory Systems Using MOF Materials and AI Capable of Distinguishing Tens of Thousands of Odors and Gases
Expanding Applications Expected in Disease Diagnosis, Environmental Monitoring, Smart Agriculture, and Robotic Chemical Detection

The future direction of next-generation "electronic nose" technology, capable of distinguishing tens of thousands of odors by mimicking the principles of human olfaction, has been presented. By combining metal-organic frameworks (MOF) with artificial intelligence (AI), a blueprint has emerged for an intelligent artificial olfactory system that can be used in applications ranging from disease diagnosis and environmental monitoring to industrial safety.

On July 9, Daegu Gyeongbuk Institute of Science and Technology (DGIST) announced that the research team led by Professor Hyuk-Jun Kwon from the Department of Electrical Engineering and Computer Science had published a review paper summarizing the key research trends and future development strategies for electronic nose technology utilizing MOF.

Conceptual diagram of artificial olfaction combining MOF sensor library (pure MOF, complexes, derivatives) and machine learning. Provided by the research team

The electronic nose is an artificial olfactory system in which multiple sensors respond to odor molecules, generating signals that are then learned and analyzed by AI. While there is a high potential for use in fields such as food quality control, detection of hazardous gases, environmental pollution monitoring, and disease diagnosis, commercialization has been restricted due to limitations in selectivity, response speed, and operating environments of conventional sensors.

Implementing 'Combinatorial Coding' Modeled After Human Olfaction

The research team focused on MOF as a key material to overcome these limitations. MOF is a porous material formed by combining metal ions with organic compounds, effectively adsorbing odor molecules through its fine pores. Its structure and chemical properties can be freely designed, making it a next-generation sensor material that can sensitively detect a variety of odors at room temperature with low power consumption.

The core of this study is the incorporation of the human olfactory recognition principle into the design of the electronic nose. Humans can distinguish tens of thousands of odors with only hundreds of olfactory receptors, thanks to "combinatorial coding," where a single odor stimulates multiple receptors simultaneously to produce a unique response pattern.

The research team proposed a technical strategy for implementing this principle by arranging MOF sensors with different properties and analyzing their signal patterns using AI.

Combining AI for Applications from Disease Diagnosis to Robotics

The team systematically categorized MOF-based electronic noses into MOF, MOF-complexes, and MOF-derivatives, detailing the advantages and application potential of each category. They particularly explained that integrating machine learning and deep learning allows for even more accurate classification and interpretation of complex odor signals.

Research team of Professor Hyuk-Joon Kwon, Department of Electrical Engineering and Computer Science, DGIST. (From right) Professor Hyuk-Joon Kwon, Hyung-Tae Lim, integrated master's and doctoral student. Provided by DGIST

Professor Kwon stated, "MOF provides a virtually limitless material library that can be designed to respond differently to a variety of odors, much like human olfactory receptors. This paper is significant in that it connects material development with AI-based odor recognition research and presents a roadmap for the development of intelligent electronic noses tailored for specific applications."

The research team expects that, in the future, MOF-based electronic noses will expand their range of applications to include healthcare technologies for disease diagnosis through breath analysis, air quality and industrial safety monitoring, smart agriculture, and chemical detection technologies for autonomous vehicles and robots.

This research was led by Hyung-Tae Lim, an integrated master's and Ph.D. student, as the first author, and Professor Kwon as the corresponding author. The results were published in 'Progress in Materials Science,' a leading international journal in the field of materials science (IF 42.9, JCR top 0.7%).

This content was produced with the assistance of AI translation services.