Until now robotics systems have found application primarily in highly structured environments, for example as manipulators in an assembly line, where robotic tasks are highly repetitive and can be largely preprogrammed and the environment is carefully controlled. In the few instances where robotic systems are operated outside of the factory, they usually rely on close human supervision.
However, recent breakthroughs in decision making, perception architectures, and mechanical design, among others, are paving the way for autonomous robotic systems carrying out a wide range of tasks of unprecedented complexity—think of autonomous space vehicles, drones, self-driving cars, and unmanned underwater vehicles.
The goal of this session was to provide a representative overview of the recent algorithmic and mechanical advances that are enabling the design and deployment of robotic systems where autonomy is pushed to the extreme, resulting in exciting innovation that borders on science fiction. Specifically, the session highlighted breakthroughs at the interface of advanced decision making and bioinspired mechanical design that are enabling first-of-a-kind applications of autonomy in space (pinpoint landing of space rockets), in air (design of micro unmanned aerial vehicles), on land (high-performance legged robotic systems), and in water (autonomous underwater vehicles).
The first speaker, Lars Blackmore from Space Exploration Technologies (SpaceX), started off by discussing autonomy in space. He is the coinventor of the G-FOLD algorithm for precision landing on Mars, and his team recently completed the first precision landing of a booster stage. He discussed his work on the autonomous precision landing technology for the Grasshopper and F9R-Dev rockets.
Next, David Lentink from Stanford University discussed autonomous, bioinspired micro flying robots. His innovations are revolutionizing the design of these robots, and he presented the ideas that made it possible.1
The session’s third speaker, Sangbae Kim from MIT, addressed autonomy on land. He and other researchers at MIT have created the robotic cheetah, “the first four-legged robot to run and jump over obstacles autonomously.” He explained how this robot is able to manage highly dynamic activities such as balance, energy, and impact without human interaction.1
Finally, Derek Paley, from the University of Maryland, looked at autonomy under water, specifically his work on motion guidance for ocean sampling by underwater vehicles.
1 Papers not included in this volume.