Karen Feigh is an associate professor in the Daniel Guggenheim School of Aerospace Engineering. Feigh is active in the design of cognitive work support systems for individuals and teams in dynamic sociotechnical settings, including airline operations, air transportation systems, and unmanned aerial vehicle ground control stations.
Feigh recently developed an interface to assist astronauts with safely and efficiently executing extravehicular activities. She has also focused on better understanding how people can naturally and intuitively teach machines to perform tasks. “Imagine that you owned a robot to help perform household chores, but the robot didn’t know how you liked your clothes folded or where to put them, what dishes needed to be handwashed instead of going into the dishwasher, et cetera,” she illustrates. “In these types of situations, the integration of robotics into our work processes requires that machine learning agents are readily able to learn from human teachers and to do so in very straightforward and nonfrustrating ways.”
“In the context of urban and regional air mobility, there have been ongoing discussions as to whether these aircraft will be autonomous and/or if we can decrease the amount of pilot training required. That opens up a wide range of research questions that we need to explore. For example, will we have autonomous algorithms that can detect if a pilot is doing something that appears to be unsafe and, if so, when should those algorithms trigger action? Will we have emergency pilots who are on the ground somewhere and, if so, how can we design systems for them to handle a workload that will likely involve highly congested air space?”
More information about Feigh’s work may be found here.