Data-centric artificial olfactory system based on the eigengraph

The research team of Professor Seong Chan Jun of the Department of Mechanical Engineering (first author, Master Seung-Hyun Sung, contributing author Yun Ji Hwang), the research team of Professor Ho Won Jang of the Department of Materials Science and Engineering at Seoul National University (co-first author, PhD Jun Min Suh), and Jeon Gue Park, CEO of Tutorus Labs Inc. have conducted joint research on the topic of 'Data-centric artificial olfactory system based on the eigengraph'. In this paper, they developed and integrated sensor arrays, electrochemical measurement devices and artificial intelligence algorithms that correspond to the functions of the olfactory receptors and each region of the brain responsible for the human olfactory mechanism to implement a data-centric artificial olfactory system and its standard model was presented. The law of existence of the eigengraph in electrochemistry was proven and declared for the first time in the world by capturing the unique electrochemical reaction that naturally exists between sensing materials and gas molecules. Based on the waveform analysis principle for eigengraphs, they have shown that various types and mixed states of gas molecules can be identified with nearly 100% accuracy even in all weather and unknown environments by extracting odor characteristics necessary for the identification of gas molecules through mathematical algorithms. They explained that this is the result of overcoming the limitations of existing sensitivity-dependent artificial olfactory technology that is vulnerable to external environmental changes. This technology can be applied and used in a wide range of fields, including artificial intelligence robots, space exploration, home appliance industry, defense industry, and forensic investigation. This study was published in a world-renowned international academic journal 'Nature Communications' (Impact Factor: 17.694).

The link: doi.org/10.1038/s41467-024-45430-9