Skip to Content

Using Bees to Detect Bombs

Honeybees might one day join the front line of national security.
December 7, 2006

Last week, scientists at Los Alamos National Laboratory, in New Mexico, buzzed with the results of a rigorous study on sophisticated bomb detectors. Their research suggests that contained bees can be used to identify volatile compounds like TATP, the primary charge associated with last summer’s terrorist plot. Highly reliable and precise, these next-wave detectors are cheap to produce and easy to train.

Los Alamos scientists look to honeybees in their quest to build a better bomb detector.

Entomologists have long known that honeybees can be trained to detect many scents, including the olfactory footprints of deadly explosives. This latest research reinforces those findings and suggests an approach that could prove useful for finding substances in populated areas.

Timothy Haarmann, principal investigator of the Los Alamos project (officially called the Stealthy Insect Sensor Project), says he and his colleagues trained bees to extend their proboscises–tubular organs used to suck the nectar from flowers–in the presence of explosives. When the proboscis is extended, the bee appears to be sticking out its tongue.

Training 50 bees requires only two or three hours using this traditional approach, which takes advantage of an insect’s attraction to sugar water. “If you hold up sugar water [to bees], they stick out their proboscis,” Haarmann says.

By combining a target substance with sugar water and then presenting the compound to the bee, the researchers manipulate the insects into recognizing a distinct smell. By the end of the session, successfully trained bees extend their proboscises toward explosives.

Bees trained at one concentration of vapor easily recognize lower doses. Chemist Robert Wingo, who works on the project, says that the bees proved to be more sensitive than many sophisticated man-made devices. “They are capable of detecting TATP, and the instruments I have available in the lab are not able to detect TATP,” he says.

Honeybees can also pick explosives out of more complicated bouquets–like the myriad scents that surround a typical human being. Trained bees can identify explosives whose odors were masked by “lotions, underarm deodorants, and tobacco products,” Wingo says. “Much to our surprise, the bees are capable of picking out TNT in motor oil … Even in the presence of insect repellent, we can train them to detect TNT.”

In Haarmann’s system the bees are contained in tubes so that their proboscises can be easily monitored. Unfortunately, a contained bee only lasts about two days. “We find that after about 48 hours you start to get a high mortality rate,” Haarmann says. Being confined is “hard on them.” Plus, not all bees prove to be up to the task of detecting explosives. Like dogs, some of the insects are more successfully trained than others. “We like to think of bees as these nice little robots, but there were certain bees that did better than others,” Haarmann says.

Jerry Bromenshenk, a researcher with the University of Montana’s division of biological sciences, is one of the pioneers of bee detection systems. He has trained bee colonies to detect explosives, meth labs, and dead bodies, but he uses a different approach. Bromenshenk works primarily with free-flying bees that are allowed to roam large, outdoor spaces. When the bees detect the target scent, they tend to slow down and circle the area. Using audio, video, and laser systems, Bromenshenk and colleagues can analyze the flight patterns of thousands of trained bees and produce a density map indicating the most likely locations of the target substance. With tens of thousands of bees searching, they can quickly canvass an area of a mile.

But Bromenshenk says Haarmann’s “bee in a box” approach still has its place.

“Free-flying bees don’t work well in airports,” he says.


Jim Tumlinson, an entomologist and the director of the Center for Chemical Ecology, at Pennsylvania State University, says the biggest barrier to enlisting real insects as cutting-edge sensors is “finding some practical way to use this behavior.”

Tumlinson, who did not work on the Los Alamos project, has researched the biomechanisms of boll weevils and parasitic wasps. He sees more potential in a biomemetic approach, in which researchers gain inspiration from nature to develop man-made systems. Currently, he is working with a multidisciplinary team of scientists developing a mechanical sensor that mimics the operation of an insect’s antenna.

Moving from the laboratory to the real world can introduce complicated obstacles, he says. “If you’re in the laboratory, you can get these insects to respond fairly reliably,” Tumlinson says. “In any field situation, the conditions are hard to control. It gets much more difficult.” But, he says, the high sensitivity of these insects is too fascinating to ignore.

“It’s very tempting to think we can do something with it, and maybe we can,” he says. “We’re in the process of learning as much as we can about how natural systems operate.”

Haarmann and his team carried out field trials, and he believes that a bee-driven bomb detector may be only a year away. He envisions remotely controlled robots in battlefields, capable of carrying a small army of honeybees to a suspected IED (improvised explosive device) or car bomb. If the bees stick out their tongue, a bomb is close by.

“You lose a couple bees, and that’s disturbing to me,” says Haarmann, who keeps his own hives and used to teach beekeeping in South America. “But I’m the only one who is disturbed.”

Wingo, who had never worked with bees prior to this project, estimates that he received “hundreds” of stings during the 18-month research-gathering period. “It’s proven to be extraordinarily interesting,” he says, “but being stung is not fun.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.