IFLScience By Janet Fang April 20, 2015
By digitally reconstructing the complete brain of the European honeybee, Apis mellifera, researchers with the Green Brain Project hope to one day create an autonomous flying robot that thinks, senses, and acts like the sophisticated pollinator.
"Bees and all other insects are miracles of engineering which we are nowhere near equaling," University of Sheffield’s James Marshall tells BBC. "If we could even recreate a fraction of their abilities in a robot system then we would have made a tremendous advance." The honeybee has surprisingly advanced cognitive behaviors, despite how simple and small their brains are compared to that of vertebrates. They can do so much with so little.
So far, the team—comprised of researchers from the Universities of Sheffield and Sussex—have recreated the parts of the bee’s brain that allows them to see and smell. Not only will this help us better understand both bee brains and human brains alike, but the team have also managed to upload their computer simulation (complete with thousands of virtual neurons) to an unmanned aerial vehicle.
To see if their bee brain simulation can actually pilot an aircraft, they plugged their simulation into a quadcopter drone and allowed it to fly down a corridor. It ended up doing so without even running into anything, Phys.org explains, and another time, it was able recognize the checkerboard pattern on the wall and use it to help navigate. This video from last November is the first demo of the quadcopter using its bee vision to navigate:
Several teams around the world are working on bee-inspired robots. Harvard researchers, for example, have a designed a RoboBee that might help pollinate fields of crops—something that might become necessary if populations of honeybees continue to dwindle. One day, perhaps different teams could come together, placing simulated bee brains inside of bee bots.
With UAVs expected to perform various dangerous missions—ranging from search-and-rescue to monitoring nuclear power plants to wildfire surveillance—it’s going to be increasingly important to control them effectively.