Remote controlled cockroaches on the march

Cockroaches have been transformed into remote-controlled living robots by a team of US scientists who sent electrical impulses to the nerves that control their legs.

Hong Liang and her colleagues at Texas A&M University have made one of the most successful attempts yet to integrate robotics with living organisms. These hybrid systems might one day become scouts for search and rescue teams, sent scurrying through collapsed buildings and disaster zones searching for survivors.

Insect are the perfect platform for such cyborg systems – and not only because they come ready assembled by nature, says Hirotaka Sato, who works on hybrid robotic beetles at Singapore’s Nanyang Technological University. “They require much less energy to move than normal robots,” he says. Another advantage is that the hybrid robots’ natural instincts mean they can take care of themselves, scurry away from danger and even refuel by eating.

One of the first cyborg insects was created in 1997, when Isao Shimoyama from the University of Tokyo successfully implanted impulse-emitting electrodes into cockroach antennae. The cockroaches perceived electrical impulses in their right antenna as a looming obstruction on the right so would swerve left, and vice versa. Like a pair of electric reins, the electrodes allowed the researchers to steer a cockroach in the direction they wanted. It all seemed promising, until the cockroaches figured out these obstructions in their path were mere phantoms, and stopped responding to the impulses.

Liang’s team found a better way to remotely control cockroaches, they report in Journal of The Royal Society Interface. Instead of targeting antennae, they embedded electrodes into ganglia, clusters of nerve fibres down the centre of the roaches’ bodies that control the insects’ limbs.  

Liang has high hopes for her little cyborg roaches.

To remotely steer a cockroach, the researchers strapped a backpack containing a small circuit board, a rechargeable battery and copper wire electrodes to the insect’s back. An electrode was then inserted into each side of the nerve cluster that controlled the insect’s front legs.

By pressing buttons on a controller, they sent electrical impulses to the electrodes buried in the roaches’ bodies. Stimulating one side of the nerve cluster triggered the leg on the same side to push out sideways, which propelled the insect in the opposite direction. The degree of control over the cockroaches could be fine-tuned even more by changing the frequency and voltage of the electrical input, which altered the sharpness of the turns. Upping the voltage could even cause the cockroach to stop moving completely.

With this level of control, the researchers didn’t quite have the cockroaches performing tap-dance routines. But they did make the cockroaches go where they wanted them to 60% of the time. As the trials went on, though, the cockroaches’ reactions grew sluggish. Was this a case of the cockroaches acclimatising to the electrical input, just as they had with the antenna signals?

No, says Liang. The responsiveness of the cockroaches was restored with a short rest. After two hours of running weighed down by a relatively hefty backpack, the cockroaches had become tired. Liang plans to make the backpacks smaller.

As for the rest of the failures, Liang chalks them up to mechanical issues, such as the electrodes sliding out of place as the cockroaches move about or the battery depleting. There is also the issue of simple human error, as anyone who’s ever tried to pilot a remote controlled car could probably attest to.

The team might have greater success if they could stimulate individual nerve cells, rather than a whole cluster of nerves, says Sato: “If there was a way to implant tiny electrodes into specific nerves, you could probably induce motion at a much higher success rate.”

Still, Liang has high hopes for her little cyborg roaches. “There’s a reason why cockroaches are one of the oldest species on the planet; they can eat almost anything, regrow their limbs and even survive nuclear radiation,” says Liang. She imagines them going where no human or regular robot can – perhaps being assigned to intelligence-gathering missions or search and rescue operations, and even nuclear clean-up situations.

Could this technology be used to help people with spinal cord injuries? The same principle could definitely be applied to trigger movement in other animals, says Liang, but you’d need a precise map and detailed understanding of their nervous systems to know how and which nerves to stimulate – and not all animals are as simple as a cockroach.