This project addresses power and control questions central to achieving maneuverable, autonomous, untethered, insect-scale, flying robots. The amount of energy stored per unit mass in even the best small batteries is low enough that flight times would be limited to at most a few minutes. Furthermore, when the high rate at which the battery must supply energy is considered, it is questionable whether a winged microrobot could even lift its own weight. Instead, this project emulates flying insects in nature, which use animal fat as fuel, with an energy density about 100 times greater than state-of-the-art batteries. Accordingly, this project will demonstrate liquid-fuel catalytic combustion engines expected to be capable of over 90 minutes of powered flight. This result will be achieved through innovative integrated modeling, analysis, design, fabrication, and control of insect-scale aerodynamics, combustion, and flight. Broader impacts will arise from application of insect-scale flying robots to, for example, artificial pollination, search-and-rescue operations, and field biological research. Indirectly, this project will produce new methods for energy conversion, novel algorithms for control synthesis, and fabrication techniques, applicable to a wide gamut of wheeled, winged, and legged microrobots.
To accomplish the project goals, the research advances knowledge in three specific areas: (i) biologically inspired design and fabrication of aerodynamically efficient flapping-wing microflyers, where principles from nature are translated into robotic designs, employing a systems-and-control conceptual framework. In this framework, the interaction between aerodynamics, power, design, and controls is analyzed using tools such as input-output modeling and system identification; (ii) mechanical actuation using fuel-powered shape-memory-alloys-based mechanisms, where flameless catalytic combustion generates the heat required to induce material phase transitions, necessary for the production of mechanical work; (iii) control, which emerges naturally from the first two areas as new aerodynamically efficient designs require the invention of novel control strategies for stable flight and new techniques for controller synthesis. Similarly, new actuation technologies require the invention of new low-level, physically implementable controllers. In this case, the dynamics of the combustion-driven actuators and flapping mechanisms are nonlinear and time-varying, reason for which a significant part of the research effort is dedicated to the development and real-time implementation of novel robustly stable nonlinear and adaptive controllers.
