Cornell develops robotic insect

Flying insects can perform impressive acrobatic feats, simultaneously sensing and avoiding a striking hand or landing on moving surfaces, such as leaves or flowers blowing in the wind. Similarly, walking insects can display amazing speed, maneuverability, and robustness by rapidly sensing and avoiding predators, while foraging or seeking shelter in small spaces and unstructured terrains.

Silvia Ferrari, Sibley School of Mechanical and Aerospace Engineering, with Robert J. Wood (Harvard University), is working toward a future where autonomous, small-scale robots would have similar capabilities, sensing and responding to their environments and maneuvering without human commands. These robots would be particularly invaluable for surveillance or reconnaissance missions in dangerous or remote environments.

Agile maneuvers require fast sensors with high accuracy and low latency, which typically translates into more processing and battery power as well as greater weight. Ferrari and Wood are overcoming this bottleneck by developing integrated sensorimotor processing, planning, and control methods that would allow fully autonomous insect-inspired robots to carry out multiple tasks with speed and maneuverability, like their biological counterparts. Ferrari and Wood are developing event-based programming methods for: cost-effective and fast multimodal sensory integration and navigation; multiple, coordinated functionalities; and robust response to disturbances.

Robots at this scale have unique advantages, such as decreased cost, covertness, physical robustness, and access to unstructured and narrow spaces inaccessible to humans. By developing new and more effective sensorimotor architectures applicable at the gram or sub-gram scale, the project is making an important leap toward the fabrication of fully autonomous small-scale robots.

Source: Cornell Research