Over his 30-year career, Bruno Thuillier has helped launch and scale some of the world’s most innovative telecommunication products, mobile phones and sensors. Before co-founding BOYDSense, Bruno was the Chief Technology Officer at Sensitive Object. He also served as Vice President of R&D at Siemens Mobile Phones, and studied at the Conservatoire National des Arts et Metiers. Now he acts as CTO of Alpha MOS and BOYDSense.
Explain us concretely how Alpha Mos AI solutions work? What needs do your solutions solve?
Alpha MOS sensory measurement solutions based on Artificial Intelligence are designed to measure and quantify product characteristics from human sensory perceptions, such as sight, smell or taste. These solutions are commonly used in the food, cosmetics and pharmaceutical industries to help assess the quality and sensory properties of products.
The operation of these solutions can vary depending on the applications and needs of our customers but generally include four distinct phases:
Why can the sense of smell be seen as complicated to grasp from an AI perspective? What are some of the challenges you face?
Smell is complicated to understand from an AI perspective for several reasons.
First, odors are complex chemical stimuli that are perceived by the human olfactory system. There are thousands of different volatile organic compounds that can be mixed in the air to create an odor. Each volatile organic compound interacts with specific odorant receptors in the nose, creating a unique neural response in the brain. This complexity makes it very difficult to create an artificial olfactory sensor that can detect and quantify odors.
In addition, the sense of smell is very subjective and varies from person to person. Olfactory perceptions can also vary according to context, age, health and other individual factors. This makes it difficult to define a universal reference model for odor identification.
Finally, it is difficult to replicate the environmental conditions necessary to detect odors accurately. The concentration of volatile organic compounds, air humidity, temperature, and other factors can influence odor perception.
To address these challenges, our researchers are working to create artificial odor sensors that are capable of detecting complex mixtures of volatile organic compounds, objectively characterizing odors, and mimicking the environmental conditions in which odors are detected. Artificial intelligence models can also be trained to recognize odor patterns based on data from artificial olfactory sensors, as well as data from sensory evaluations by humans. This approach can help establish reference patterns for odor identification and improve the accuracy of olfactory detection.
According to you, what will be the next revolution in AI?
It is difficult to predict with certainty what the next AI revolution in sensory analysis will be, as advances in this field are happening rapidly. The next revolution in sensory intelligence, which refers to the ability of machines and systems to process and interpret information from sensory sensors, could include the following developments: