Job Talk: Interpersonal Autopilots: Developing Autonomous Drones which are Interactive, Intuitive, and Ethical
Apr 13, 2016
from 1:25 PM to 2:15 PM
|Where||135 Reber Building, University Park, PA 16802|
|Contact Name||Michelle Barnyak|
|Add event to calendar||
Unmanned aerial systems (UASs) are currently being proposed which will utilize a wide array of autonomous functionality including military targeting, emergency landing, and operating with and in close proximity to people. The next generation of autopilots will be expected to model the people around the vehicle and to take their goals, behavior and the context into consideration while making mission-related decisions. Moreover, these systems will need to intuitively communicate with non-experts, allowing people, for example, to calibrate their trust to the actual performance of the system. The focus of this presentation is on human-robot trust, an important topic for a wide range of real world UAS tasks ranging from military to search and rescue and humani- tarian missions. Methods from game theory and social psychology are used to develop representations, algorithms and a system architecture which allows an autonomous system to predict a person’s trust in the system, whether the sys- tem should trust a person, and to quantify their trust in terms of risk. These abilities are demonstrated in live scenarios in which a robot provides guidance to a person during an emergency evacuation showing that the design of a robot can easily influence people to trust the robot too much. These findings suggest that cognitive modeling of people by an autonomous system will be important if we are to create systems which will safely operate in and among the public.
Alan Wagner is a Senior Research Scientist at the Georgia Tech Research In- stitute. Wagner’s research interests include social robotics, human-robot inter- action, cognitive modeling, game theory, trust and deception. His work focus- es on real world problems related to search and rescue, manned-unmanned teaming, and situation awareness. His work on human-robot trust during emergency evacuations was recently featured in New Scientist, The Christian Science Monitor, Popular Mechanics, Discover, and on NPR. Dr. Wagner’s research has been recognized by the human-robot interaction community win- ning a best paper award at RO-MAN 2007 and a best paper finalist in 2014, an Air Force Young Investigator Award, and Georgia Tech Research Institute’s Innovation Award. Wagner has contributed to the organization to several workshops and conferences related to human-robot interaction and artificial intelligence. He earned his BA in Psychology from Northwestern University, a Master’s degree in Computer Science from Boston University, and his doc- toral degree in Computer Science from Georgia Institute of Technology under the supervision of Professor Ronald Arkin, a Machine Ethics pioneer.
About the Speaker
Senior Research Scientist Georgia Institute of Technology