Scientists created a wheeled robot that can smell with locust antennae

The robot's biological sensor sensitivity to smell is 10,000 times higher than current electronic devices.
Deena Theresa
The robot with the biological sensor.
The robot with the biological sensor.

Tel Aviv University 

In a scientific first, researchers at Tel Aviv University built a robot that can "smell" using a biological sensor - the locust antennae. According to the makers, the robot's biological sensor sensitivity to smell is 10,000 times higher than current electronic devices.

The sensor sends electrical signals in response to a nearby odor, which can be detected and interpreted by the robot as per a release.

The researchers connected the biological sensor to an electronic system first. Using a machine learning algorithm, the robot could identify odors with a high level of sensitivity. According to the team, this technology could be pretty helpful in the future in identifying explosives, drugs, diseases, and more.

The results of the study were published in the journal Biosensor and Bioelectronics.

Detecting each odor at the level of the insect's primary sensory organ

According to the researchers — doctoral student Neta Shvil of Tel Aviv University’s Sagol School of Neuroscience, Dr. Ben Maoz of the Fleischman Faculty of Engineering and the Sagol School of Neuroscience, and Prof. Yossi Yovel and Prof. Amir Ayali of the School of Zoology and the Sagol School of Neuroscience — our sensory organs, such as the eye, ear, and nose, and well as those of animals, use receptors that identify and distinguish between different signals. 

The sensory organ then translates these findings into electrical signals that are decoded as information.

"We connected the biological sensor and let it smell different odors while we measured the electrical activity that each odor induced. The system allowed us to detect each odor at the level of the insect’s primary sensory organ," Prof. Yovel said in a statement.

Machine learning created a library of smells

In the second step, the team used machine learning to create a ‘library’ of smells. They could characterize eight odors, such as geranium, lemon, and marzipan, in a way that allowed them to distinguish between smells. 

"In fact, after the experiment was over, we continued to identify additional different and unusual smells, such as various types of Scotch whiskey. A comparison with standard measuring devices showed that the sensitivity of the insect’s nose in our system is about 10,000 times higher than the devices that are in use today," Prof. Yovel said.

"The principle we have demonstrated can be used and applied to other senses, such as sight and touch. For example, some animals have amazing abilities to detect explosives or drugs; the creation of a robot with a biological nose could help us preserve human life and identify criminals in a way that is not possible today," said Dr. Maoz.

In the future, the researchers plan to guide the robot toward navigation, allowing it to localize the odor source and, later, its identity.

Study Abstract:

Identifying chemical odors rapidly and accurately is critical in a variety of fields. Due to the limited human sense of smell, much effort has been dedicated to the development of electronic sensing devices. Despite some recent progress, such devices are still no match for the capabilities of biological (animal) olfactory sensors, which are light, robust, versatile, and sensitive. Consequently, scientists are turning to a new approach: Bio-Hybrid sensors. These sensors combine animal biological sensors with electronic components to achieve maximum detection and classification while conveying a comprehensible signal to the end user.

In this work, we created a bio-hybrid odor discriminator utilizing the desert locust's primary olfactory apparatus - its antennae, together with simple electroantennogram technology and artificial intelligence tools for signal analysis. Our discriminator is able to differentiate between at least eight pure odors and two mixtures of different odorants, independently of odorant concentration. With four orders of magnitude higher sensitivity than gas chromatography–mass spectrometry, it is able to detect the presence of less than 1 ng of volatile compounds and, compared to other bio-hybrid sensors available today, it can be easily operated by an unskilled individual. This study thus opens up the future for robust and simple bio-hybrid robotic sensing devices that can be widely deployed.