Introduction to Human-Robot Interaction (HRI)

Transcription

1 Introduction to Human-Robot Interaction (HRI) By: Anqi Xu COMP-417 Friday November 8 th, 2013

2 What is Human-Robot Interaction? Field of study dedicated to understanding, designing, and evaluating robotic systems for use by, or with, humans [Goodrich 2007] Rethink Robotics MIT Media Lab NASA Osaka U.

3 Brief History (of HRI as a Field) 1992: 1 st IEEE International Symposium on Robots and Human Interactive Communications (RO-MAN) Late 1990 s: workshops and conference tracks on HRI at International Robotics Conferences (e.g. AAAI Symposia Series, ICRA, IROS, RSS, Human Factors, etc.) 2000: 1 st IEEE/RAS International Conference on Humanoid Robots (HUMANOIDS) 2004: IEEE/RAS & IFRR summer school on Human-Robot Interaction 2006: 1 st ACM International Conference on Human-Robot Interaction (HRI) 2012: inaugural volume of Journal of Human-Robot Interaction (jhri)

4 Related Fields Telerobotics & Haptics Human Factors Air Traffic Control Human-Computer Interaction Operations Research Machine Learning

5 Unique Features of HRI Physical interaction with embodied intelligence NASA Robonaut Rich social relations between humans and robots Fox Complex, dynamic & unpredictable environment USAR NIST / Texas U&M

6 Taxonomies in HRI Physical vs. cognitive Co-located vs. remote Team configurations [Yanco 2002] H H H H H R R R R R R H H H H H H H H R R R R R R R

7 Constituents of HRI Objective: understand and shape interactions between human(s) and robot(s) Multidisciplinary: social sciences, natural sciences, engineering, 5 Main Constituents of HRI Level and behavior of autonomy Nature of information exchange Structure of human-robot team Training of people and robots Design and shaping of task for human-robot collaborations

8 Autonomy Mapping of inputs from the environment into actuator movements, representational schemas, etc. [Goodrich 2007] Neglect Time [Goodrich 2003] Turing test Means to support productive HRI Laundry-Folding Robot (UC Berkley)

9 Autonomy (cont.) Levels of Autonomy (LOA) [Sheridan 1978] 1. Computer offers no assistance; human does it all 2. Computer offers a complete set of action alternatives 3. Computer narrows the selection down to a few choices 4. Computer suggests a single action 5. Computer executes that action if human approves 6. Computer allows the human limited time to veto before automatic execution 7. Computer executes automatically then necessarily informs the human 8. Computer informs human after automatic execution only if human asks 9. Computer informs human after execution only if it decides to 10. Computer decides everything and acts autonomously, ignoring the human direct control dynamic autonomy

10 Information Exchange Intelligent interaction requires deliberate communication Interaction Time [Goodrich 2003] Switch attention to current task Establish context Plan actions Communicate plan to robot Workload Situational Awareness Shared Mental Model

11 Information Exchange (cont.) Methods of communication: Medium Robot-Initiated Human-Initiated Visual Audio Tactile Visual displays (lights, GUI, VR, AR) Natural language, non-speech audio Haptics Gestures (hand, facial, body) Natural language Keyboard, mouse, gamepad, haptics HRP-4C (NIST) Tag-based Comm. with Aqua Robot (McGIll U.) Paro Theraputic Robot

12 Teams Motivations & Purposes Task division Task efficiency Redundancy Multi-robot management Cognitive overload Relates to level of autonomy, nature of task, and mode of communication Fan-Out [Goodrich 2003] Kestrel Autopilot v3.0 (Lookheed Martin)

13 Teams (cont.) Team hierarchy Conflict resolution Especially for peer-based relationships (co-x) Active roles Supervisor Operator Mechanic / Assistant Peer Slave Passive roles Patients Visitors Bystanders Lovotics Robot (N. Taipei U.) RIBA: Robot for Interactive Body Assistance (RTC)

14 Training Training robots (learning) Learning from Demonstration (LfD): transfer of task-domain knowledge from human teacher to robot student Training operators Learning curve Intuitive design: facilitates interaction with challenged users Necessity of training: streamline interaction, reduce risk Training designers Human-centric interface design Humans and Automation: Use, Misuse, Disuse, Abuse [Parasuraman 1997]

15 Task Shaping Fundamental changes to nature of tasks Optimize human/robot workload distribution Encourage synergy between capabilities of humans and robots Additional support to human-robot team Design tools to facilitate interaction for both human and robot Employ technology / other robots for greater situational awareness Heterogeneous Multi-Robot Coral Reef Inspection (McGill U.)

16 HRI Constituents: Summary Framing the main constituents and objectives of HRI research Level and behavior of autonomy Nature of information exchange Structure of human-robot team Training of people and robots Design and shaping of task for human-robot collaborations Dynamic interaction: time-varying, task-varying, context-specific changes to HRI constituents

17 Sample HRI Problems / Solutions Establishing design principles for efficient human-robot interaction [Goodrich 2003] Characterizing human s situational awareness [Endsley 1988] Teaching robot helicopter acrobatic moves [Ng 2004]

18 Seven Principles of Efficient Human Robot Interaction [Goodrich 2003] 1. Implicitly switch interface and autonomy modes 2. Let robot use natural human cues 3. Manipulate the world instead of the robot 4. Manipulate relationship between the robot and the world 5. Let people manipulate presented information 6. Externalize memory Raizlabs 7. Help people manage attention Virtual Cockpit v2 (Procerus / Lookheed Martin)

19 Situational Awareness Global Assessment Technique (SAGAT) [Endsley 1988] Evaluation procedure: Assign task scenario to user and robot (in simulation) At some random point in time, halt simulation and blank relevant displays Administer random subset of questions about SA requirements SA levels: immediate, intermediate, long-range Compare real and perceived situation post-hoc, and report true/false % Repeat for different users to obtain measures of statistical significance

20 Autonomous Inverted Helicopter Flight via Reinforcement Learning [Ng 2004]

21 Further Resources Conferences: HRI, RO-MAN, ICRA, IROS, RSS Journals: jhf, jhri Website: My Contact: /

22 References M. Goodrich and A. Schultz. Human-Robot Interaction: A Survey. Foundations and Trends in Human- Computer Interaction, 1(3), 2007, pp M. Goodrich and D. Olsen. Seven Principles of Efficient Interaction. in Proceedings of IEEE International Conference on Systems, Man, and Cybernetics, 2003, pp T. Sheridan and W. Verplank, Human and Computer Control for Undersea Teleoperators. MIT Man-Machine Systems Laboratory, H. Yanco and J. Drury. A Taxonomy of Human-Robot Interaction. in Proceedings of the AAAI Fall Symposium on Human-Robot Interaction, 2002, pp R. Parasuraman and V. Riley. Humans and Automation: Use, Misuse, Disuse, Abuse. Human Factors: The Journal of Human Factors and Ergonomics, 39(2), 1997, pp K. Lakshmanan, A. Sachdev, Z. Xie, D. Berenson, K. Goldberg, P. Abbeel. A Constraint-Aware Motion Planning Algorithm for Robotic Folding of Clothes. in proceedings of the 13th International Symposium on Experimental Robotics (ISER), F. Shkurti, A. Xu, M. Meghjani, J. Gamboa, Y. Girdhar, P. Giguere, B. Dey, J. Li, A. Kalmbach, C. Prahacs, K. Turgeon, I. Rekleitis, and G. Dudek. Multi-Domain Monitoring of Marine Environments using a Heterogeneous Robot Team, in Proceedings of the 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '12), Portugal, October M. Endsley. Situation Awareness Global Assessment Technique (SAGAT). In Proceedings of the National Aerospace and Electronics Conference (NAECON), 1988, pp A. Ng, A. Coates, M. Diel, V. Ganapathi, J. Schulte, B. Tse, E. Berger, and E. Liang. Inverted Autonomous Helicopter Flight via Reinforcement Learning. in International Symposium on Experimental Robotics, 2004.