Why Is MIT Making Robot Insects?

MIT’s micro-robotics push is less about building “cool insect copies” and more about solving a stack of physics and engineering problems that only show up when machines shrink to bee-size and below. At that scale, air drag, surface tension, power delivery, and even battery packaging behave differently—so the path to flight, swimming, jumping, and controlled motion requires insect-inspired mechanisms and unconventional power sources. The payoff is practical: robots that can inspect engines, search disaster rubble, and potentially operate in low-gravity environments—without needing the bulky, failure-prone designs that have historically hampered rescue efforts.

A central theme is how scale changes the rules of motion. Small flyers have far higher surface-area-to-volume ratios, which increases drag and reduces the inertia that lets larger animals glide. That’s why bees and other insects flap rapidly: wing motion generates vortices and low-pressure zones above the wing while higher pressure below provides lift. MIT’s RoboBees also borrow aerodynamic tricks from nature—maple tree seeds that spin while falling. Their shape creates swirling vortices near the leading edge, and adding miniature electric rotors at the wingtips can provide enough lift to transition from falling to powered flight.

But shrinking creates new bottlenecks. Surface tension traps tiny robots at the water’s surface, making “just swim” impossible. MIT demonstrates multiple escape routes. One yellow submarine-like robot flaps wings to move on land and underwater, but when it’s stuck at the surface, it splits water into hydrogen and oxygen, stores the gases in a buoyancy chamber, then ignites them so an explosion breaks the surface layer and launches the robot upward before it can fly. Another design uses water-repellent copper pads: applying 600 volts charges the pads to attract water molecules, breaking the hydrophobic barrier so the robot can sink on command and then walk underwater.

Power and durability are equally decisive. Early RoboBees used piezoelectric crystals to drive wing flapping, but the crystals were fragile—impacts could crack them and stop the robot. MIT’s newer approach replaces piezos with soft polymer “muscles” coated with carbon nanotubes that act like tiny actuators when voltage cycles pull and release the polymer. These muscles can stretch up to 25% of their length, flap around the 400 hertz range, and keep working through bumps and scrapes. If pierced, the carbon nanotubes can short; MIT addresses this with self-healing by burning off the shorted nanotubes and even a laser-assisted clearing process to isolate defects.

For energy limits, MIT also highlights a jumping strategy: hopping can conserve power compared with continuous flight. Separately, a cockroach-inspired inspection robot called HAMR uses voltage-controlled adhesion to stick to metal surfaces and move through tight spaces, aiming to detect turbine cracks that are otherwise expensive and slow to inspect.

Finally, MIT tackles the battery problem with a “tiny combustion engine” concept: a penny-sized unit runs on a continuous stream of methane and oxygen, ignites it, and uses expanding hot gases to push a flexible polymer membrane like a piston. The design relies on heat loss at small scales so flames don’t travel back through fuel lines. With enough thrust for jumping and carrying significant payload relative to mass, these micro-engines point toward longer-lasting robots that can carry sensors and communication hardware—while the lab keeps the emphasis on fundamental curiosity as much as on applications.