Researchers at Tel Aviv University have created what may be the world’s most sensitive odor-identification device by marrying cutting-edge electronics and AI with “technology” going back hundreds of millions of years: the locust antenna. (Relax, dear reader, the locust survives this story.)
Their proof-of-concept invention of a bio-mechanical smell-robot based on a locust’s feeler was described last week in the prestigious journal Biosensors and Bioelectronics.
- These Gummy Bears Taste Delicious. They Are Made From Locust
- First bees and ants, now people: An evolutionary transition might be coming for humanity
- ‘A Rare Treat’: Paleontologists Detect 50-million-year-old Fossil Bug Penis
“Smell-bots” are used at the more advanced airports to detect contraband such as marijuana in our luggage and on our person. Even more advanced airports use dogs because, as Prof. Yossi Yovel, Dr. Ben Maoz and the team point out, man-made technologies still can’t compete with millions of years of evolution. They still can't, but now the scientists at Tel Aviv University have taken the concept further by harnessing the advanced olfactory sense of the locust.
Insects began to emerge almost half a billion years ago and have had time to develop extraordinary abilities. Our friend the mosquito can detect a 0.01 percent difference in the concentration of carbon dioxide, which helps guide them to potential meals: animals exhaling CO2. Humankind is nowhere near producing artificial sensors with such capabilities, the team says.
Why choose locusts and not, say, mosquitoes or cockroaches, who don’t have many fans howling about their rights? Locusts were known to be able to identify a wide range of scents, says Maoz, adding that co-author Prof. Amir Ayala – an expert on the bug – thought it the appropriate starting point.
So, the Israeli desert locust smells using its antennae. The trick is connecting that biological sensor to an artificial system that can receive and decode the electrical signals – which the team calls “simple electro-antennogram technology” armed with odor-analyzing software. Their bio-mech smell-bot is 10,000 times more sensitive than a commonly used electronic “smelling” device, the team says.
What is simple electro-antennogram technology? “Information in the body of all living organisms is transferred by way of electrical impulses. Our sensory organs transduce different kinds of environmental stimuli (for example, odor molecules) into electrical signals,” Maoz explains. “Electro-antennogram technology allows the signals generated by the insect antenna in response to specific olfactory signals (specific odorants) to be ‘read.’ These signals can then be fed to a machine learning algorithm for identification and to a robot for localization.”
Teaching the machine involved connecting the antenna to the technology and exposing it to various odors, while measuring the bio-electrical activity that each odor induced, Yovel says.
“The system allowed us to detect each odor at the level of the insect’s primary sensory organ. Then, in the second step, we used machine learning to create a ‘library’ of smells,” he explains.
Thusly they were able to characterize the antenna’s perception of eight odors, such as geranium, lemon and marzipan. “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,” Yovel notes.
The resultant device can distinguish between at least those eight pure odors and two mixtures of different odorants, irrespective of odorant concentration, the team says.
The ability to identify eight odors sounds unhelpful. But that number is just an artifact of the bug-bot’s training process, Maoz explains.
“Most electronic sniffers are sensitive to specific molecules (which means to specific odors),” he says. “Since we’re using a biological nose, we’re (almost) not limited to specific molecules. We were able to see unique responses to dozens of odors.” But the eight is because training the system to recognize a specific scent – which odor belongs to which material – is time-consuming. One can train the sniffer to detect any odor one wishes and in this case they chose eight. Marzipan smugglers, you stand warned.
The locust made it
Now one wonders what happened to the locust test subjects. Asked about that, Maoz explains that they detach one of the insect’s two antennae and that it can live fine with just one.
That is well and good, but a snag comes to mind regarding future product development of a hybrid bio-mech sniffer-bot. Organic material decays fast. How long can a detached locust antenna last?
“For this study, we created a device that can prolong the antenna’s functionality for up to 12 hours,” Maoz says. “This was sufficient for this work, but we may develop additional methods that will prolong it more if needed.”
In other future work, the researchers say, they plan to give the smell-bot navigational ability so it can trundle up to the odor sources.
Many of us know the airport detection devices: this author has a fond memory of being significantly delayed at JFK decades ago because of a substance in the suitcase that crashed the airport’s brand-new smell-bot (it was asafoetida). But odor-identification technology has much broader uses, including detecting explosives, cancer and more.
It is very early stages to discuss commercialization, but the team has big visions: Maoz says they are already in discussions with various governmental as well as commercial entities.
“Nature is much more advanced than we are, so we should use it,” he says. “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. Some animals know how to detect diseases ... the sky’s the limit.”
The project was led by doctoral student Neta Shvil of Tel Aviv University’s Sagol School of Neuroscience, Maoz, Yovel and Ayali of the School of Zoology and the Sagol School of Neuroscience.