At the Okinawa Institute of Science and Technology in Japan, neurobiologist Kenji Doya is ankle-deep in rodents. Not real ones, but “cyber rodents” made of plastic and silicon. Two of the critters circle each other in a mating dance. Others forage for fresh batteries on the floor. Another one just sits there. “That one is lazy,” says Doya, who also heads a group at the Advanced Telecommunications Research Institute in Kyoto. “It doesn’t expend energy to get a reward” – and probably won’t last long.
Groups of robots have been fixtures in academic robotics labs for years. But Doya’s project is one of the first to use robots to probe how administering rewards to individuals when they achieve simple goals can give rise to intelligent group behaviors. This work could help designers build machines that collaborate to carry out complex tasks. By studying how groups of mobile robots interact and adapt, researchers could eventually develop self-sufficient swarms of robots that explore hostile environments, gather surveillance data, and repair equipment remotely.
The key, says Doya, is teaching the robots to do the right thing. Each 22-centimeter-long cyber rodent is equipped with a processor chip, a camera, sensors, motorized wheels, and infrared data ports that allow it to communicate, or “mate,” with others. If a robo rat approaches a battery pack or orients itself to mate, it receives a digital “reward” – a snippet of software code that reinforces that behavior in the future. Over time, says Doya, the robots compete for power and may even develop territories and alliances.
“It’s early-stage but very promising,” says Terry Sejnowski, a computational neuroscientist at the Salk Institute for Biological Studies in La Jolla, CA. “Kenji’s robots have a clever algorithm to develop sophisticated behaviors.”
It could be years before such robots do useful work outside the lab. But Junku Yuh, program director of robotics and computer vision at the National Science Foundation, says funding multirobot systems is important because they could lead to more efficient ways to control machines and gather information in the field. To that end, researchers at the University of Pennsylvania, the Georgia Institute of Technology, and Burlington, MA–based iRobot are developing new theories about and military applications for swarms of robots. There is safety in numbers – but there could also be smarts.
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Deep learning pioneer Geoffrey Hinton has quit Google
Hinton will be speaking at EmTech Digital on Wednesday.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.