The University of Alabama
February 1, 2017
Robotic systems are becoming more practical in military applications. In fact, unmanned aerial vehicles (UAVs) are currently used for surveillance and reconnaissance missions. However, current systems utilize complex one (or multiple) operator/one robot interfaces. Also, implementation of human-in-the-loop systems creates issues because human operators intervene more frequently if they do not trust the system or their expectations of the autonomy are not met. As a result, excessive or inept human intervention negatively affects workload, situational awareness, and performance.
This research is aimed at allowing a single operator to efficiently manage multiple UAVs and interact effectively with high levels of autonomy. Our research considers both changes to autonomous algorithm design as well as novel updates to the user interface. Autonomous algorithms are updated to use a more natural quality of service metric, Average Linear Uncovered Length (ALUL), to better match operator mental model in an effort to increase trust. In addition, the user interface is updated to include additional exocentric cues to update operator understanding of the autonomous actions selected by the system. Simulations that study the performance of the new algorithms show that surveillance performance is comparable for unconstrained systems and enhanced for constrained systems. A human subject study using our UAV surveillance simulation system compares manual and semiautonomous surveillance performance using training and additional interface information to increase trust and teaming. Results show providing a UAV-based egocentric view of autonomy can enhance that operator teaming with multiple UAVs.
Monica Anderson is an Associate Professor in the Computer Science department at The University of Alabama. Her research focuses on distributed autonomous systems that enable search, rescue and reconnaissance applications in unknown, complex environments. Other research interests include smart environment technologies that support aging in place and disabled persons through personalized autonomy and safety management. She is also dedicated to making both robotics and computer science more accessible through novel programming interfaces and updated frameworks.
Dr. Monica Anderson graduated with honors from Chicago State University in 1990 with her BS in Computer Science and from the University of Minnesota in 2007 with a PhD in Computer Science and Engineering. She has authored over 45 publications and has served as Pi or Co-PI on grants in excess of $1.5 million dollars. She is an active member of IEEE and ACM and in 2008, received the UPE Excellence in Instruction award. Dr Anderson is a CoPI of the NSF-funded BPC Alliance Institute for African American Mentoring in Computing (iAAMCS). iAAMCS focuses on increasing the number of African Americans in Computer Science earning graduate degrees through novel interventions. Between earning her BS and PhD, Dr. Anderson worked as a software engineer. Her career included working on point-of-sale and stores systems for Target, architecting web-based banking applications for Norwest Bank (now Wells Fargo) and designing distributed travel assistance applications for Northwest Airlines (now Delta Airlines).