Some trained dogs risk life or limb in their job to detect drugs or even explosives with their powerful noses, but they may soon be replaced by robots that can already perceive scents.
Researchers at Tel Aviv University managed to give a robot the sense of smell using a biological sensor from an insect. The sensor sends electrical signals in response to the presence of a nearby scent, which the robot can detect and interpret.
The researchers managed to connect the biological sensor to an electronic system and, using a machine learning algorithm, were able to identify odors with a level of sensitivity 10,000 times higher than that of a commonly used electronic device.
The scientists who developed the robot claim that “the sky is the limit”, and believe that this technology could also be used in the future to identify explosives, drugs, diseases, etc.
The researchers point out that, in general, our sensory organs such as the eye, ear and nose – as well as those of all other animals – use receptors that identify and distinguish between different signals. The sensory organ then translates these findings into electrical signals, which the brain decodes as information.
“We connected the biological sensor (to the electronics) and let it smell different odors while we measured the electrical activity that induced each odor. The system allowed us to detect each odor at the level of the insect’s primary sensory organ.”
— Yossi Yovel, a professor at Tel Aviv University’s School of Zoology and the Sagol School of Neuroscience.
Thus, the challenge of biosensors, such as the one used in the robot in question, lies in connecting a sensory organ, such as the nose, to an electronic system that knows how to decode the electrical signals received from the receptors.
“Nature is much more advanced than we are, so we need to take advantage of that,” says Dr. Ben Maoz, a professor at the Fleischman School of Engineering and the Sagol School of Neuroscience at Tel Aviv University and an author of the study.
The authors explain that they connected the biological sensor (to the electronics) and let it smell different odors while they measured the electrical activity that induced each odor. The system allowed them to detect each scent at the level of the insect’s primary sensory organ.
After getting the system to detect the scents, the developers used machine learning to create a “library” of scents.
Scientists hope that this technology can be used to preserve human lives by detecting explosives and to identify criminals by identifying drugs.
“Some animals have amazing abilities to detect explosives or drugs; creating a robot with a biological nose could help us preserve human lives and identify criminals in a way that is not possible today,” Maoz concluded.
10,000
times more sensitive than current electronic reference devices (known as mass spectrometers) is the robot.
How can this robot smell?
-The robot is a biohybrid system
-It is made up of a biological component, which is the antenna of a lobster (the organ that smells) and an electronic component, which is the robot itself.
-The biological antenna creates a unique electrical response to each odorant.
-This electrical signal is “interpreted” by the robot and reacts accordingly.
Interview
Dr. Ben Maoz,
professor at the Fleischman School of Engineering and the Sagol School of Neuroscience at Tel Aviv University
Q: What is behind the idea of giving a robot the sense of smell?
– The general idea was that our current technology cannot compete with some of the sensory capabilities of the insect kingdom. For example, smell. Currently there are no good sensors for odors, unlike in the case of insects.
Our thinking was that instead of trying to mimic these sensory capabilities with electronic systems, we would create biohybrid systems, where we would integrate the insect’s sensory organ (for example, the antenna) with a robot, which would respond to the biological sensors.
The idea is a perfect example of the importance of holding face-to-face meetings (as opposed to online meetings). Yossi Yovel, a biosensor and bat expert, Amir Ayali, a locust expert, and I met for coffee, and after brainstorming, we came up with this idea.
Q: How does it work?
– The odorant identification process consists of several stages. First, the electrical signal generated by the antenna is “cleaned” and processed, so that it does not have background noise. The next step is to use the AI to train the system on what the unique characteristics of each odorant are. Once trained, the system can identify and discriminate these odorants.
Q: What applications can this technology have?
– The sky is the limit. It can be used for many things, such as identifying explosives, drugs, food, perfumes, diseases and anything with a smell that you want to identify and distinguish.
Q: How sensitive is this robot to odors?
– In the measurements we made, we found that it is about 10,000 times more sensitive than current reference electronic devices (known as mass spectrometers). Keep in mind that it depends on the odorant and the insect that we are using.