Eight weeks ago, local sev­enth and eighth graders in an after-​​school sci­ence club got an up-​​close look at real sci­en­tific research while vis­iting Northeastern’s Marine Sci­ence Center in Nahant, Mass. The field trip, attended by stu­dents from the Buck­ingham Browne & Nichols School in Cam­bridge, Mass., com­menced a semester-​​long cur­riculum designed by North­eastern doc­toral can­di­date Daniel Blus­tein as part of his effort to teach youth how to trans­late the orga­ni­za­tional beauty of nature into con­trol sys­tems for robotic devices.

I want to inspire stu­dents to get into sci­ence because I think sci­ence is awe­some,” said Blus­tein, who spends most of his time designing robotic lob­sters and bees in biology pro­fessor Joseph Ayers’ lab. The Bio­mimetic Under­water Robot Pro­gram takes its own inspi­ra­tion from the neural net­works of real ani­mals to design more adapt­able robots, he said.

Blus­tein piloted the Neu­ro­bots cur­riculum two years ago with the stu­dents’ teacher, Kelly Schultheis, who holds a master’s degree in biology from Northeastern.

Eighth grade BB&N stu­dents Max Wie­gand and Ava Boudreau work on their robotic lob­ster before the race. Photo by Brooks Canaday.

The pro­gram con­sists of two sec­tions. First, the stu­dents learn some basic animal biology. They prac­tice observing organ­isms in their nat­ural habi­tats by taking objec­tive notes about the ani­mals’ behavior, and they learn about neural net­works by playing games in which the stu­dents pre­tend to be neu­rons passing a signal (in the form of mar­bles) from one person to the next.

In the second half of the pro­gram, they begin to learn basic com­puter pro­gram­ming and robot design. Using Lego Mind­storms as the plat­form, the stu­dents design a robot that must maneuver through a simple maze. The trick, how­ever, is that its com­mands don’t come from a remote con­troller, but rather from the envi­ron­ment around it—just as with real animals.

Blus­tein has devel­oped a visual pro­gram­ming lan­guage to approx­i­mate neural net­works that inter­faces with the stan­dard Lego NXT soft­ware. The stu­dents can drag and drop bits of this visual code on the com­puter to tell their robots to do things like moving right if they encounter a wall on the left or moving back­ward if they encounter a wall up ahead.

This is how most ani­mals work, one of the middle school stu­dents said during the final club meeting in which they raced their robots. Cues on one side of the body cause a phys­ical response on the oppo­site side. Lob­sters col­lect this infor­ma­tion with their antennae and claws, but the Lego robots do it with sensors.

Some stu­dents designed their robots to respond to the walls using touch sen­sors. Others used light sen­sors to track the black tape run­ning along the perimeter of the maze. Still others used a com­bi­na­tion of the two.

Kelly Schultheis, left, and Clint Valen­tine a third-​​year biology and envi­ron­mental sci­ence major, right, look on as 7th and 8th grade sci­ence club stu­dents from BB&N watch a competitor’s robot nav­i­gate the maze. Photo by Brooks Canaday.

In the end, the Lego robots encoun­tered some sig­nif­i­cant chal­lenges as they made their way through the maze. One got caught in a spin­ning loop, or “doing a jig,” as its designer put it. Another was “doing the right thing, it’s just doing it back­wards!” exclaimed a student.

Despite the set­backs, a pal­pable excite­ment charged the room as the stu­dents cheered on each other’s robots and anx­iously watched as the victor crossed the finish line.

Blustein’s own interest in marine life began with reg­ular child­hood trips to the National Aquarium in Bal­ti­more. He later became inter­ested in research while working with octo­puses at the Seattle Aquarium during col­lege. “Sci­ence isn’t com­plete until you form some kind of com­mu­ni­ca­tion with the public,” he said. “You need to extend what you do, to share it with others.”