ced Decision Architectures Collaborative Technology Alliance Five Lessons Learned in Human-Robot Interaction Patricia McDermott Alion Science and Technology Jennifer Riley, Ph.D. SA Technologies Douglas Gillan, Ph.D. North Carolina State University Laurel Allender, Ph.D., Army Research Laboratory
23
Embed
Advanced Decision Architectures Collaborative Technology Alliance Five Lessons Learned in Human-Robot Interaction Patricia McDermott Alion Science and.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
• Lesson 5: Removing the human from the environment to be navigated disrupts the perception of visual information about the environment and the motion. The HRI interface can be designed to reduce this disruption and enhance the operator’s ability to navigate a robot through its distant environment and to perform the desired tasks.
– Widen the field of view (FOV) if possible. Providing a third-person perspective also benefits navigation, the benefits are additive.
– Temporal resolution may be sacrificed if the resolution is high and the robot is traveling at relatively low speeds. A high spatial resolution aids in target detection and identification.
Lesson 4: When robotic operators are distributed from Soldiers who need the information from robotic assets to complete a mission, the interfaces and collaborative technologies must support the communication of high quality spatial information in relation to the Soldier in the field.
• The following solutions can be implemented to aid collaboration among humans with robotic tasks:– When robotics team members cannot be co-located, it is
advantageous to incorporate aspects of co-location such as the ability to share imagery.
– When sharing imagery, be sure to include enough reference information so that the recipient understands where the picture was taken in reference to his or her location.
– Allow both team members to have full-time access to a map of the area without having to search between displays.
– Allowing team members to communicate visually on a shared map can enhance the communication of highly spatial information.
Lesson 3: Operating a robot inherently involves multitasking and task switching. When designing automation to aid multitasking, designers must consider how social and human factors in HRI can interact to negatively counter the benefits of automated functions.
• When implementing automation in robotic tasks, designers should keep in mind the following:– At natural break points in tasks or during down time, the
system could prompt participants to switch tasks or robots. This would facilitate appropriate task switching strategies that produce better performance without sacrificing SA
– Provide critical cues to help operators reconcile imagery from different perspectives such as aerial and ground or from two different ground perspectives. Shape and color were the most salient cues followed by texture, markings, and shadows.
– Evaluate the coordination costs of two operators sharing one automated asset. The costs may outweigh the benefits, especially in an ad-hoc team.
Lesson 2: Factors associated with human control of robotics assets can differentially impact SA. It is critical to determine the information needs of Soldiers within a context and to support the multiple facets of SA through system and interface design.
In order to provide information at the right level, start with a goal-directed (cognitive) task analysis
• Support for visually demanding tasks. Examples include object recognition support, strategies or interface support for depth perception and judging size and distance, and support for parallel processing and divided attention.
• Support for understanding robot localization and situated-ness. Examples include improved GPS for robots, mapping tools at the interface, interface designs that present data on orientation and robot configuration, and mission recording and playback for quick SA update.
• Improved interface design for developing higher level SA. For example, conduct SA-oriented design of interfaces to facilitate integration of data at the interface and design for understanding of robot automation states and modes).
• Ways in which humans and robots can work seamlessly together in naturalistic environments
• Implications for display design, interaction design, team configuration, collaborative technologies, and HRI metrics
• Future research – Development of system design requirements– Examination of interrelationships in HRI, i.e., spatial
awareness and SA, spatial awareness and mental map development
– More studies should be done on the interrelationships among human performance measures such as performance time and accuracy, SA, teamwork, and collaboration