How Artificial Intelligence is Revolutionizing Olfactory Technology
Summary:
AI now enables computers to detect scents, marking a groundbreaking advancement in technology.
(AIM)—Over time, artificial intelligence (AI) has transformed the way we interact with technology. Today, it even allows computers to detect smells, which is truly remarkable. More than a century ago, Alexander Graham Bell encouraged National Geographic readers to measure odors. He noted that while the sciences of sound and light were well-advanced, the science of smell remained unexplored. This is why he called for the creation of a new scientific discipline. Today, with AI beginning to penetrate the field of olfaction, Bell’s call still resonates. Thanks to AI, computers have learned to perceive.
Our smartphones are equipped with facial and voice recognition, but it wasn’t until recently that they gained olfactory capabilities. This change is due to machine olfaction or digitized smell. Despite significant challenges, progress is being made in this area. The complexity of human smell presents enormous challenges. In fact, while vision relies on a few types of cells in the retina, olfaction uses about 400 different types of receptors. Olfactory machines start with sensors that detect molecules in the air, mimicking the receptors in the human nose. However, for this technology to be useful, it must identify and interpret these molecules like humans do.
The Key Role of Machine Learning
Deep learning, a specific type of machine learning, plays a crucial role in these advancements. It allows machines to link the molecular structure of compounds to olfactory descriptors. In other words, the model learns to associate words like “sweet” or “floral” with specific molecules. This requires extensive datasets, but historically, olfactory data has been scarce. The situation began to change in 2015 with the DREAM Olfaction Prediction Challenge. Led by biologists Keller and Vosshall, this competition shared olfactory data for teams to make predictions. The best-performing results were published in the journal Science in 2017. A technique known as “random forests” was particularly effective.
Practical Possibilities of Machine Olfaction
The advancements in machine olfaction are not merely theoretical; they have significant practical implications. For instance, they enable the creation of personalized perfumes or the development of better insect repellents. Additionally, early disease detection through the analysis of body odor becomes a real possibility. Digitized olfaction can also improve personal hygiene by making individuals more aware of their odors.
Expanding Datasets and the Impact of COVID-19
Datasets have dramatically expanded, especially after the COVID-19 pandemic highlighted olfactory disorders such as anosmia. For example, the Pyrfume project has made a vast amount of olfactory data publicly available. From less than 500 molecules in 2019, the dataset grew to about 5,000. This has allowed teams like Alexander Wiltschko’s at Google to advance deep learning in this field. Today, Wiltschko runs Osmo, a company dedicated to bringing the sense of smell to computers. These advancements suggest that the future of machine olfaction is not only promising but also exciting.
The integration of AI in olfaction marks a significant leap in technology. By teaching computers to smell, AI is opening new frontiers in personalized products, medical diagnostics, and beyond. As datasets expand and technology improves, the potential applications of machine olfaction continue to grow, making this a truly groundbreaking field.
Follow us on Facebook: AI Insight Media.
Get updates on Twitter: @AI Insight Media.
Explore AI INSIGHT MEDIA (AIM): www.aiinsightmedia.com.
Keywords:
AI smell technology, machine olfaction, digitized smell, deep learning, olfactory data, personalized perfumes, early disease detection, COVID-19, Pyrfume project, Osmo
