If AI has a sense of smell, what can it do?

As a start-up that uses artificial intelligence and digital olfactory technology to simulate human olfaction, Aryballe is helping corporate customers to transform odor data into truly actionable information.

Many friends may not remember the practical effects of such a technology for a while. To this end, we interviewed Sam Guillaume, CEO of Aryballe, and asked him about the realization principle of digital olfaction, its application in the current market, and his personal predictions about the future of olfactory technology.

How does the nose tell the smell?

Our nose processes the odor molecules released by organic and inorganic substances to make distinctions and judgments. When the energy in the object increases (including pressure, agitation or temperature changes), the odor molecules will naturally escape, and we can inhale it through the nasal cavity and produce a corresponding olfactory experience.

Odor molecules stimulate our nasal olfactory neurons and olfactory bulbs. Our brain combines other information (such as visual cues and our past memories of smells) to accurately recognize the current smell and make follow-up judgments based on this.

The realization principle of digital smell

Digital smell first uses biosensors to capture odor characteristics, and then uses software solutions to analyze and display odor data to simulate human response to odors as much as possible. Artificial Intelligence (AI) interprets and classifies the signatures based on the previously collected smell database.

Guillaume said, “In the past few years, we have fundamentally replicated the realization of human olfaction through technical means. For example, we transplanted this technology to existing technologies such as semiconductors to create small, convenient and easy-to-use technologies. Use and low-cost sensors. Our results are very close to human beings in terms of odor discrimination ability and effectiveness."