Deploying highly maneuverable vehicles autonomously in real-world settings represent one of today's main challenge in guidance and control. Autonomous flight is very limited in adaptability and agility when compared to piloted flight. This disparity is particularly striking for small-scale aircraft. Miniature aircraft can exhibit maneuvering capabilities far exceeding those of traditional ones, which would potentially allow them to negotiate intricate environments and perform complex tasks. The talk will illustrate this chasm and provide an overview of the challenges and accomplishments in autonomous flight.
I will first provide some insight into the physical origins of high maneuverability for miniature rotorcraft type vehicles. I will describe the application of identification modeling techniques for the development of dynamic models and the extraction of performance characteristics. I will use this data for the analysis of the effects of scale on some key performance and maneuvering characteristics. I will conclude this part by briefly describing human piloting modalities observed during flight testing and the development of a control logic that enables the automatic implementation of extreme maneuvers.
In the second part of my presentation I will outline the salient characteristics of the problem of autonomous flight and describe our choice of methodologies. In particular I will give an overview of the dynamic models developed for real-time trajectory planning and I will describe the disparities between the short- and long-term planning requirements. I will present work aimed at integrating both domains in an approaches aimed at satisfying the computational and functional requirements. Finally, the challenge of autonomous flight involves many other problem areas. To conclude my presentation I will give a short overview of research opportunities that are tied with my work.