Panelists: Moderator Amy Pritchett, professor and head, Department of Aerospace Engineering, Pennsylvania State University; James Cutler, associate professor of aerospace engineering, University of Michigan; Aaron Kahn, flight controls engineer, U.S. Naval Research Lab; Fritz Langford, program manager, Aurora Flight Sciences
Michele McDonald, AIAA communications manager
AIAA SCITECH FORUM, San Diego, Jan. 10, 2019 — Panelists in the “Autonomy Across Domains”
session discussed how autonomy is working behind the scenes in vehicles as disparate as space-bound rockets, Vietnam-era military helicopters and balsawood gliders.
“The magic here is hard to show,” said Amy Pritchett, professor and head of the
Department of Aerospace Engineering at Pennsylvania State University.
Autonomy can accomplish tasks that would otherwise be impossible, said James Cutler, associate professor of aerospace engineering at the
University of Michigan. For example, autonomous systems can handle landing in a dark crater on an asteroid, which can’t be mapped in advance, or can control vehicles on Mars, despite an approximate 20-minute communications delay to Earth.
“You can’t joystick or control anything that far away,” Cutler said.
There are more than 80 programs and 400-plus vehicles at the
U.S. Naval Research Lab, said Aaron Kahn, NRL flight controls engineer. The autonomous vehicles are designed to be deployed by ship, air or air to water.
One such vehicle is a fully autonomous, electric coaxial helicopter that launches from a tube and looks like a flying fence post. It has 10 seconds to figure out which way is up, unfold and become an aircraft. Another is a glider, and like Buzz Lightyear, “it falls with style” Kahn said.
Participants in the panel discussion "Autonomy Across Domains,” Jan. 10 at the 2019 AIAA Science and Technology Forum and Exposition (AIAA SciTech 2019) in San Diego.
The utility military helicopter Bell UH-1 Iroquois, aka Huey, was transformed into an autonomous vehicle and tested last year, said Fritz Langford, program manager at Aurora Flight Sciences, a Boeing company. The autonomous Huey is designed to be controlled by a human working with a tablet. In one test, the helicopter landed in the hazardous conditions of a brown-out, or large dust cloud.
Challenges to autonomy include environment, how to handle failure, figuring out what an operator intends without overloading them with options, and learning how to trust a machine. Machine learning could be the next big breakthrough for aerospace, panelists said.
Where humans fit into the autonomous design is another challenge. If people are expected to take over from the machine when something goes wrong, they need time to assess the situation, and that time needs to be designed into the system, Langford said.
Cutler said autonomous systems come with moral questions, including what to automate and why.
“How does this autonomy help us to be human?” Cutler asked.
Watch the full video for more details.
Return to 2019 AIAA SciTech Forum Headlines
Return to 2019 AIAA SciTech Forum Home