de
en
Schliessen
Detailsuche
Bibliotheken
Projekt
Impressum
Datenschutz
zum Inhalt
Detailsuche
Schnellsuche:
OK
Ergebnisliste
Titel
Titel
Inhalt
Inhalt
Seite
Seite
Im Dokument suchen
Meyer zu Borgsen, Sebastian: Nonverbal Communication During Human-Robot Object Handover. Improving Predictability of Humanoid Robots by Gaze and Gestures in Close Interaction. 2020
Inhalt
Abstract
Acknowledgments
Contents
List of Research Questions
List of Hypotheses
List of Requirements
List of Figures
List of Tables
Research Topic
1 Introduction
1.1 Outline
2 Thesis Prospect
2.1 Research Questions
2.2 Research Hypotheses
2.3 System Requirements
2.3.1 Handover Structure
2.3.2 Predictable and Natural Motions
2.3.3 Interaction with Everyone
2.3.4 Required Perceptional Capabilities
2.3.5 Design and Integration
2.3.6 Summary
3 Terminology and Concepts
3.1 Robot
3.2 Human-Robot Interaction
3.3 Nonverbal Communication
3.3.1 Gestures as Information Channel
3.3.2 Gaze for Mutual Understanding
3.4 Object Handover
3.4.1 Classification of Handovers
3.4.2 Phases of Handover
3.4.3 Grasping for Handover as a Giver
3.4.4 Positioning for Handover
3.4.5 The Object Transfer Point
3.4.6 Reaching Out
3.4.6.1 Considering Only the End-Effector Motion
3.4.6.2 Full Joint Trajectory
3.4.7 Transferring Physical Control
3.4.8 Laterality and Handedness
3.5 Summary
The Object Handing Robot
4 System Foundation and Environment
4.1 Meka M1 Mobile Manipulator
4.2 System Architecture
4.3 Reproducibility and Software Deployment
4.4 Environment of the Robot and Experiments
4.5 Incorporating Mixed Reality
4.5.1 Mixed Reality Human-Robot Interactions
4.5.2 Augmenting the User's Perspective
4.5.3 Simulating Parts of the Robot
4.5.4 Simulating the whole Robot
4.5.5 Mixed Reality Summary
5 Human-Robot Handover Experiment
5.1 Experiment Procedure and Design
5.1.1 Robot Behavior
5.1.2 Gestures with the Second Arm
5.1.3 Basic Gaze Strategy
5.1.4 Setup and Environment
5.2 Experiment Annotation
5.3 Findings and Results
5.3.1 Timing Analysis
5.3.2 Influence of Gestures
5.3.3 Influence of Experience
5.3.4 Discussion
6 Using Robot Gaze for Predictability
6.1 Floka – The Interactive Head
6.2 Aligning Human-Like Gaze with Handover
6.2.1 Aligning to Handover Phases
6.2.2 Timing and Duration
6.3 Gaze Integration with Floka
6.3.1 Input Stimuli
6.3.2 Comparison with Non-Interactive Gaze
7 Functional Gesture Motions
7.1 Combining Recorded and Dynamic Motions
7.1.1 Handover Motion Database
7.1.2 Adaptive Motions
7.2 Integration with Floka
7.2.1 Reaction to Input
7.2.2 Blending of Trajectories
7.2.3 Adaption Motion Generation
7.3 Resulting Motions
7.3.1 Reaching in Simulation
7.3.2 Reaching Motions on the Real Robot
8 Perception and Prediction of OTPs
8.1 Hand Tracking
8.2 Transfer Point Prediction
8.3 Integration with Floka
8.3.1 Optimal Sensor Mounting
8.3.2 Hand Tracking Pipeline
8.3.2.1 Tracking Initialization
8.3.2.2 Continuous Tracking
8.3.2.3 Tracking Recovery
8.3.3 OTP Prediction
8.4 Tracking and Prediction Results
9 Combined Handover System
9.1 The Final Floka Robot
9.2 Combined Behavior
9.2.1 Enhanced Gaze
9.2.2 Approaching - Orient Towards Interactant
9.2.3 Reaching Behavior
9.2.3.1 Prediction and Adaption Interaction
9.2.3.2 Improved Transfer Detection
9.2.4 Object Transfer
9.2.5 Retreat
9.3 Parameters
9.4 Discussion
10 Final Evaluation
10.1 Interaction Design and Procedure
10.1.1 Study Setup
10.1.2 Handover Interaction Tasks
10.2 Appraisal
10.3 Discussion
10.4 Summary
Perspectives
11 Discussion & Outlook
11.1 Current Limitations
11.2 Outlook and Future Work
12 Conclusion
Appendix
A Gesture Study Material
A.1 Recorded ROS-Topics
A.2 Handover Questionnaire 2017
A.3 Instructions
A.3.1 Versuchsleiter Anweisungen
A.3.2 Examiner Instructions [translated]
B Gesture Study Timing Plots
B.1 Duration by Experience
B.2 Duration by Condition
C Adaption Plots
D Final System-Evaluation Material
D.1 Recorded ROS-Topics
D.2 Behavior Statemachine
D.3 Handover Questionnaire 2019
D.4 Handover Instructions 2019
D.4.1 Interaktionsaufgaben
D.4.2 Scripted Interaction Tasks [translated]
D.5 Handover Interview 2019
D.5.1 Interview Fragen
D.5.2 Interview Questions [translated]
E Evaluation Duration as Boxplot
Acronyms and Abbreviations
Glossary
Bibliography
Own Publications
RoboCup ToBi Involvement
General References
Online Resources
Software Packages
Declaration of Authorship
Colophon