The first flight of an insect-sized robotic fly

Insects of the order Diptera have evolved to become prolific flyers able to perform aerial maneuvers that far surpass anything man-made. The Harvard Micro robotics lab has recently demonstrated the first step towards recreating these evolutionary wonders with the world´s first demonstration of an at-scale robotic insect capable of generating sufficient thrust to takeoff (with external power). The mechanics and aerodynamics of this device are quite similar to Dipteran insects. Biologists have recently quantified the complex nonlinear temporal phenomena that give insects their outstanding capabilities. Periodic wing motions consisting of a large stroke and pronation and supination about an axis parallel to the span-wise direction are characteristic of most hovering Dipteran insects. Previous microrobot designs have attempted to concisely control each wing trajectory in these two dimensions. The robot that is shown here has three degrees-of- freedom, only one of which is actuated. Here, a central power actuator drives the wing with as large a stroke as possible and passive dynamics allow the wing to rotate using flexural elements with joint stops to avoid over-rotation. There are four primary components to the mechanical system: the actuator (or ´flight muscle´), transmission (or ´thorax´), airframe (or ´exoskeleton´) and the wings. Each is constructed using a meso- scale manufacturing paradigm called Smart Composite Microstructures. This entails the use of laminated laser-micromachined materials stacked to achieve a desired compliance profile. This prototyping method is inexpensive, conceptually simple, and fast: for example, all components of the fly can be created in less than one week. Additionally, the resulting structures perform favorably when compared to alternative devices: flexure joints have almost no loss, ultra-high modulus links have higher stiffness-to-weight than any other material, and the piezoelectric actuators have similar power density to th- e best DC motors at any scale. After integration, the fly is fixed to guide wires that restrict the motion so that the fly can only move vertically. The wings are then driven open loop to achieve a large angular displacement. This is done at resonance to further amplify the wing motion. The wings exhibit a trajectory nearly identical to biological counterparts. Finally, this 60 mg, 3 cm wingspan system is allowed to freely move in the vertical direction demonstrating thrust that accelerates the fly upwards. Bench-top thrust measurements show that this robotic fly has a thrust-to- weight ratio of approximately two. These results unequivocally confirm the feasibility of insect-sized MAVs. The remaining challenges involve the development of microelectronics appropriate for power conversion, sensing, communication, and control along with the choice of an appropriate power source.