By Zak Stone,Zak Stone

Scientists at two U.K. universities are now collaborating on an attempt to create a highly accurate model of a honeybee's brain as part of a project to better understand how animals think. By reproducing the key systems that make up a bee's perception, like vision and scent, the researchers will eventually be able to install the artificial bee brain in an independently acting, flying robot, that will be able to carry out complex tasks, similar to the brain of a real animal.

"The development of an artificial brain is one of the greatest challenges in artificial intelligence," says Dr. James Marshall of the University of Sheffield, the project's lead (the other researchers are from the University of Sussex). "So far, researchers have typically studied brains such as those of rats, monkeys, and humans, but actually 'simpler' organisms such as social insects have surprisingly advanced cognitive abilities."

And because of its small size, it lends itself toward actual use in a flying robot some day, which could be used on assignments like identifying odors or gasses, assisting with search and rescue missions, or even mechanically pollinating crops.

The team has named the project "Green Brain"--a sly reference to the Blue Brain project, an initiative led by IBM that's attempting to harness the power of supercomputers to replicate the functionality of a human brain. But unlike that supercomputer cluster-enabled project, Green Brain runs on GPU accelerators, the same type of processors responsible for 3-D graphics in game consoles, and a much more efficient option.

The project comes at a time of huge interest in mechanically replicating the work of bees, which play an essential role in pollinating food crops but are on the decline. At Harvard, the Robobee project is attempting to create a mechanical bee, and although it will lack the "intelligence" of the Green Brain, researchers announced just recently that they're finally able to steer it--a big step toward their goal of using it to pollinate fields.

Comments