When a scout hon­eybee returns to the hive, she per­forms a “waggle dance,” looping and shaking her rear end in par­tic­ular pat­terns to direct her com­rades toward the jackpot of nectar and pollen she’s found. Her move­ments com­mu­ni­cate the direc­tion and dis­tance of the nectar source, pro­viding a vector along which the other bees can now travel. As they fly through the air, the flow of optical stimuli across their periph­eral vision tells the bees how far they’ve trav­eled and when to turn.

The whole oper­a­tion is the nuanced output of the bees’ neural cir­cuitry, the product of eons of evo­lu­tionary opti­miza­tion. Now that the hon­eybee pop­u­la­tion is in steep, inex­plic­able  decline, the loss of its spe­cial­ized pol­li­na­tion prac­tices threatens crop via­bility across the globe.

But if hon­ey­bees dis­ap­pear com­pletely, a team of sci­en­tists at North­eastern, Har­vard Uni­ver­sity, and CentEye, Inc. has a plan. Using Harvard’s ground­breaking pop-​​up man­u­fac­turing tech­nique, the team can rapidly gen­erate inex­pen­sive swarms of minia­ture flying robots, which could some day pol­li­nate an entire field of crops.

But, a swarm of micro-​​robots could be used for a lot of dif­ferent things,” said Dan Blus­tein, a grad­uate stu­dent at Northeastern’s Marine Sci­ence Center. They could be used for traffic and weather mon­i­toring or to safely inves­ti­gate a leak at a radi­a­tion plant, for instance.

A lot of the tech­nology of this stuff, if you can shrink it down that small, get the power require­ments that low,” said post-​​doctoral researcher Anthony West­phal, “it really does open up a lot of win­dows in terms of what can be useful.”

Blus­tein and West­phal are mem­bers of pro­fessor Joseph Ayer’s bio­mimetic research lab, which inves­ti­gates the neural net­works of var­ious animal species and repli­cates them in robotic sys­tems. The lab first began inves­ti­gating the neural net­works under­lying behavior in lob­sters and lam­prey and to date has pro­duced sev­eral gen­er­a­tions of RoboLob­sters and Robo­Lam­prey that sense and respond to their envi­ron­ment using an elec­tronic neural net­works of neu­rons and synapses that mimics the animal model’s brain. These under­water sys­tems are being devel­oped for under­water mine countermeasures.

There are many sim­i­lar­i­ties between bees and lob­sters,” said Blus­tein. While the other groups on the RoboBee team are working to master the body design and its sen­sory equip­ment, the bees wouldn’t get very far without an instinc­tive method for responding to the envi­ron­ment. Optical flow data col­lected by visual sen­sors on the bee’s head needs to be trans­lated into adap­tive move­ment in one direc­tion or another.

Most arti­fi­cial intelligence-​​based robots are con­trolled algo­rith­mi­cally,” said Ayers, the prin­ciple inves­ti­gator on the National Sci­ence Foundation-​​supported research. This means the designer must pre­dict and gen­erate com­puter pro­grams for every pos­sible con­tin­gency of the envi­ron­ment in which the robot oper­ates, he explained. Animal behavior, in con­trast, is con­trolled by neu­ronal and synaptic net­works that the team mimics in what they call “bio­log­ical intelligence.”

We are adapting con­trollers derived from animal ner­vous sys­tems to the con­trol of robotics,” said Ayers. The unique thing with the hon­ey­bees is it’s the first time anyone has attempted neuronal-​​based bio-​​mimicry with a flying plat­form. Moving in three dimen­sions intro­duces a host of new com­pli­ca­tions, which the team is now addressing.

Col­leagues at Har­vard have intro­duced a group behavior com­po­nent that will allow the RoboBees to not only to orga­nize pol­li­na­tion mis­sions, but to pass along what they’ve learned in some robotic ver­sion of the waggle dance.