In their pursuit of making artificial intelligence possess the same sensory capabilities as humans, scientists have largely focused on making AI (Artificial Intelligence) adept at dealing with visual information. But a new wave of researchers is now focusing on enabling the AI to detect and analyze smells.
The necessity of teaching AI to identify smells
The reason why AI needs to be equipped with the ability to smell (olfaction) is simple — it provides an extra set of information to make more comprehensive and accurate judgments in any situation. A good example would be that of AI-powered cars.
Researchers used to believe that a few cameras and visual information would be sufficient in developing an AI self-driving vehicle. But they soon understood that relying solely on imagery alone actually restricts AI from making accurate decisions. Since the vehicle is constantly moving through different environments, adding in a sense of smell will provide AI with more information so that it can drive safely and efficiently.
“Every type of stimulus doesn’t get processed in the same way… Vision and olfaction are very different types of signals, for example… So there may be different strategies to deal with different types of data. I think there could be a lot more lessons beyond studying how the visual system works,” Wired quotes Saket Navlakha, a computer scientist from the Salk Institute for Biological Studies, California.
An AI-powered car equipped with the capability to detect smells may be able to identify whether the region it is moving through has toxic gases present. The AI could then shut down all windows of the car and prevent the passengers from being exposed to such toxicities.
AI capable of smell
A research team from the Loughborough University is developing AI that can analyze human breath and detect hidden ailments. When completed, AI can be used in hospitals and clinics to quickly diagnose patients.
“Using this technique, deep learning systems can be trained to detect small amounts of volatile compounds with potentially wide applications in medicine, forensics, environmental analysis, and others,” Andrea Soltoggio, a lecturer at the Loughborough University, says in an article at The Conversation.
Another group of researchers from the HSE Laboratory of Space Research, Technologies, Systems, and Processes is developing an electronic nose capable of detecting and discriminating between various types of gas compounds and mixtures. The nose mimics olfactory functions of real-world organisms in that it stores the smells it encounters, thereby making it easy to identify it next time around.
“Our main objective at the moment is to increase the range of olfactory patterns the device can recognize by enabling it to learn new smells and commit this information to memory. Essentially, we want to teach the device to discriminate between hazardous and non-hazardous gas mixtures and memorize them quickly. For this purpose, it needs to know the characteristics of each gas,” Phys.Org quotes researcher Vladimir Kulagin.
The electronic nose has a wide range of application. For instance, when used in underground mines, the electronic nose will be able to warn workers when levels of toxic gases increase beyond the prescribed limit. In addition, it can also be used by security agencies to detect bombs in a specific region, thereby saving lives.