Shared Control of a Guided Parafoil and Payload System

Abstract

The direct inclusion of human pilots into airdrop operations has a strong potential to increase the landing accuracy of conventionally autonomous guided airdrop systems. Human pilots have significant mental flexibility and an innate ability to prioritize requirements to aid safe and accurate landings. Autonomous algorithms on the other hand, are generally rigid in nature and cannot handle situations not directly programmed into the software. This paper outlines the work done to develop an integrated human-machine interface that combines the flexibility of human control decisions with the powerful ability of autonomous systems to measure data on the aircraft and generate key estimates. Two interfaces are presented melding a first person view camera mounted to the payload of the airdrop system and a bird's eye GPS based map of the drop zone with relevant system estimates overlaid. Flight testing of these digital feedback methods to the pilot are studied to identify the ability of human operators to successfully and accurately control an airdrop system to the ground. Results indicated that a trained human operator has the ability to improve landing accuracy over a conventionally autonomous system by 36%.