BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:iCalendar-Ruby
BEGIN:VEVENT
CATEGORIES:Lecture or Presentation
DESCRIPTION:Decision Making for Marine Robotics\n\nUnderwater gliders\, pro
 peller-driven submersibles\, and other marine robots are increasingly being
  tasked with gathering information (e.g.\, in environmental monitoring\, of
 fshore inspection\, and coastal surveillance scenarios). However\, in most 
 existing scenarios\, human operators must carefully plan the mission to ens
 ure completion of the task. Strict human oversight not only makes such depl
 oyments expensive and time consuming but also makes some tasks impossible d
 ue to the requirement for heavy cognitive loads or reliable communication b
 etween the operator and the vehicle. We can mitigate these limitations by m
 aking the underwater robotic systems semi-autonomous\, where the human prov
 ides high-level input to the system and the robot fills in the details on h
 ow to execute the plan.  \n\nThis lecture will discuss how a general framew
 ork that unifies information theoretic optimization and robot motion planni
 ng makes semi-autonomous information gathering feasible in marine environme
 nts. The presentation will leverage techniques from multi-robot coordinatio
 n\, reinforcement learning\, and topological modeling to provide solutions 
 for (1) multi-vehicle information gathering\, (2) navigation with semantic 
 commands\, and (3) robust marine vehicle path execution. Finally\, the lect
 ure will present results from a recent deployment in the Gulf of Mexico whe
 re a team of six heterogeneous underwater and surface vehicles successfully
  tracked multiple coastal upwelling fronts over a 10-day period using the p
 roposed methods. \n\nGeoff Hollinger is an associate professor in the Colla
 borative Robotics and Intelligent Systems (CoRIS) Institute and the School 
 of Mechanical\, Industrial & Manufacturing Engineering at Oregon State Univ
 ersity. He has previously held research positions at the University of Sout
 hern California\, Intel Research\, University of Pennsylvania’s GRASP Labor
 atory\, and NASA's Marshall Space Flight Center. He received his Ph.D. (201
 0) and M.S. (2007) in Robotics from Carnegie Mellon University and his B.S.
  in General Engineering along with his B.A. in Philosophy from Swarthmore C
 ollege (2005). He is a recipient of the ONR YIP award (2017) and the NSF CA
 REER award (2019) and has served as associate editor for the Autonomous Rob
 ots and IEEE Robotics and Automation Letters journals.
DTEND:20220321T200000Z
DTSTAMP:20260511T031600Z
DTSTART:20220321T190000Z
LOCATION:
SEQUENCE:0
SUMMARY:College of Engineering Faculty Lecture: Geoff Hollinger
UID:tag:localist.com\,2008:EventInstance_42484821997316
URL:https://events.oregonstate.edu/event/college_of_engineering_faculty_lec
 ture_geoff_hollinger
END:VEVENT
BEGIN:VEVENT
CATEGORIES:Lecture or Presentation
DESCRIPTION:Decision Making for Marine Robotics\n\nUnderwater gliders\, pro
 peller-driven submersibles\, and other marine robots are increasingly being
  tasked with gathering information (e.g.\, in environmental monitoring\, of
 fshore inspection\, and coastal surveillance scenarios). However\, in most 
 existing scenarios\, human operators must carefully plan the mission to ens
 ure completion of the task. Strict human oversight not only makes such depl
 oyments expensive and time consuming but also makes some tasks impossible d
 ue to the requirement for heavy cognitive loads or reliable communication b
 etween the operator and the vehicle. We can mitigate these limitations by m
 aking the underwater robotic systems semi-autonomous\, where the human prov
 ides high-level input to the system and the robot fills in the details on h
 ow to execute the plan.  \n\nThis lecture will discuss how a general framew
 ork that unifies information theoretic optimization and robot motion planni
 ng makes semi-autonomous information gathering feasible in marine environme
 nts. The presentation will leverage techniques from multi-robot coordinatio
 n\, reinforcement learning\, and topological modeling to provide solutions 
 for (1) multi-vehicle information gathering\, (2) navigation with semantic 
 commands\, and (3) robust marine vehicle path execution. Finally\, the lect
 ure will present results from a recent deployment in the Gulf of Mexico whe
 re a team of six heterogeneous underwater and surface vehicles successfully
  tracked multiple coastal upwelling fronts over a 10-day period using the p
 roposed methods. \n\nGeoff Hollinger is an associate professor in the Colla
 borative Robotics and Intelligent Systems (CoRIS) Institute and the School 
 of Mechanical\, Industrial & Manufacturing Engineering at Oregon State Univ
 ersity. He has previously held research positions at the University of Sout
 hern California\, Intel Research\, University of Pennsylvania’s GRASP Labor
 atory\, and NASA's Marshall Space Flight Center. He received his Ph.D. (201
 0) and M.S. (2007) in Robotics from Carnegie Mellon University and his B.S.
  in General Engineering along with his B.A. in Philosophy from Swarthmore C
 ollege (2005). He is a recipient of the ONR YIP award (2017) and the NSF CA
 REER award (2019) and has served as associate editor for the Autonomous Rob
 ots and IEEE Robotics and Automation Letters journals.
DTSTAMP:20260511T031600Z
DTSTART;VALUE=DATE:20230321
LOCATION:
SEQUENCE:0
SUMMARY:College of Engineering Faculty Lecture: Geoff Hollinger
UID:tag:localist.com\,2008:EventInstance_42490885503399
URL:https://events.oregonstate.edu/event/college_of_engineering_faculty_lec
 ture_geoff_hollinger
END:VEVENT
END:VCALENDAR