跳到主要內容

簡易檢索 / 詳目顯示

回結果列表
研究生： 陳亭安
Chen, Ting-An
論文名稱： 四足機器人的恐怖谷效應：探討動靜態與擬真度對使用者親和感之影響
Exploring the Uncanny Valley in quadruped robots: how motion and realism influence user perception
指導教授： 簡士鎰
Chien, Shih-Yi
口試委員: 陳志銘
Chen, Chih-Ming
袁千雯
Yuan, Chien-Wen
學位類別： 碩士
Master
系所名稱： 商學院 - 資訊管理學系
Department of Management Information System
論文出版年： 2025
畢業學年度： 113
語文別： 英文
論文頁數： 69
中文關鍵詞： 人機互動 、四足機器人 、恐怖谷
外文關鍵詞： HRI (human-robot interaction), quadruped robot, The Uncanny Valley effect
相關次數： 點閱：19  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 本研究聚焦於Unitree Go2四足機器狗，旨在探討其不同擬真程度的外觀條件 （加予毛皮、頭套）與不同模式（靜態與行為模式）如何影響使用者的親和感及恐怖 谷體驗。透過兩階段的線上前測調查，研究分別採用修改版Godspeed問卷與HRIES、 NARS等量表來評量受試者對機器狗擬真度、親和度與不適感的感知。 結果顯示，單純增加毛皮時若搭配自然動作，使用者親和感明顯下降，呈現經 典的恐怖谷現象；然而，當加有毛皮的機器狗外觀與指令動作結合時，受試者對於擬 真度與親和度的評價分別平均提升約 0.68 分與 0.44 分，有效緩解不適感。而毛皮 加頭套的情況則不然，即使加入動態行為，也無法回到與裸機或毛皮單一條件相當的 舒適水準，顯示外觀的一致性在決定恐怖谷深度上扮演主導角色，而行為僅能作為次 要的調節機制。 依據此發現，本文提出一套針對四足機器人的外觀—行為設計框架，建議在外 觀擬真度尚屬中等程度時，可透過動態行為策略來提升使用者體驗；但當外觀不一致 度超出補償上限，仍需進一步在外型層面進行調整，以真正降低恐怖谷效應。

    This study examines the Unitree Go2 quadruped robot dog to determine how variations in its visual realism—specifically the addition of fur and headgear—and its operational mode (static versus dynamic) influence user affinity and the uncanny valley effect. In a two-stage online pilot study, participants completed a modified Godspeed questionnaire alongside the HRIES and NARS scales to rate perceived realism, affinity, and discomfort. The findings indicate that applying only fur to the robot’s body while it executes natural movements significantly reduces user affinity, producing a classic uncanny valley dip. In contrast, when the fur-only condition is paired with behaviorally congruent, commanded actions, ratings of perceived realism and affinity increase by an average of 0.68 and 0.44 points respectively, thereby alleviating discomfort. However, in the fur-plus-headgear condition, even the introduction of dynamic behaviors fails to restore comfort to the levels observed in the bare or fur-only configurations. These results underscore that morphological consistency is the primary determinant of uncanny valley depth, whereas behavior functions only as a secondary tuning mechanism. Building on these insights, we propose an appearance–behavior co-design framework for quadruped robots. When visual realism remains within a moderate, compensable range, targeted dynamic behaviors can enhance user experience and “lift” the design out of the valley. Yet, once appearance incongruence surpasses a critical threshold, deeper morphological adjustments are required to genuinely mitigate the uncanny valley effect.

    ACKNOWLEDGEMENTS 2
    ABSTRACT 4
    TABLE OF CONTENTS 5
    TABLE OF TABLES 7
    TABLE OF FIGURES 8
    CHAPTER 1: INTRODUCTION 9
    CHAPTER 2: BACKGROUND 11
    2.1 HUMAN-ROBOT ENCOUNTER AND INCOP SITUATION 11
    2.2 THE UNCANNY VALLEY IN ZOOMORPHIC ROBOTS 13
    2.2.1 Emotional/ psychological factors of the uncanny valley effect 14
    2.2.2 Measuring the Uncanny Valley effect 15
    2.2.3 Interaction Design 17
    CHAPTER 3: METHODOLOGY 19
    3.1 FIRST PILOT TEST DESIGN 19
    3.2 DATA COLLECTION OF THE FIRST PILOT TEST 20
    3.3 PRELIMINARY RESULTS OF THE FIRST PILOT TEST 21
    3.4 SECOND PILOT TEST DESIGN 24
    3.5 DATA COLLECTION OF THE SECOND PILOT TEST 26
    3.6 PRELIMINARY RESULTS OF THE SECOND PILOT TEST 27
    3.6.1 Descriptive Statistics 27
    3.6.2 Correlation Analysis 28
    3.6.3 ANOVA analysis 28
    3.6.4 Post Hoc Comparisons 29
    3.6.5 Qualitative Feedbacks 29
    CHAPTER 4: IN-LAB STUDY 30
    4.1 EXPERIMENTAL DESIGN 30
    4.1.1 Power Analysis and effect size 32
    4.2 PROCEDURE 33
    4.3 DATA COLLECTION 34
    4.3.1 Pre-test Survey 34
    4.3.2 Experiment Survey 35
    CHAPTER 5: RESULTS 37
    5.1 DESCRIPTIVE STATISTICS 37
    5.1.1 Participants 37
    5.1.2 Survey 38
    5.1.3 The Uncanny Valley 46
    5.2 DATA ANALYSIS 48
    5.2.1 Affinity for Technology Interaction 49
    5.2.2 Dog-Likeness 49
    5.2.3 Animacy 50
    5.2.4 Likeability 51
    5.2.5 Perceived Safety 51
    5.2.6 Disturbance 52
    5.2.7 NARS Sub-scales 52
    5.2.8 Uncanny Valley 54
    5.3 QUALITATIVE RESULTS 55
    5.3.1 Bare x Natural 55
    5.3.2 Bare x Trained 56
    5.3.3 Fur Only x Natural 56
    5.3.4 Fur Only x Trained 56
    5.3.5 F+H x Natural 56
    5.3.6 F+H x Trained 56
    CHAPTER 6: DISCUSSION 56
    6.1 SUMMARY OF QUANTITATIVE FINDINGS 56
    6.1.1 Behavioral Congruence as a Moderator of the Valley 56
    6.1.2 A Compensation Ceiling for Cross-Modal Inconsistency 57
    6.1.3 Appearance Dominance with Secondary Behavioral Tuning 57
    6.1.4 Expectation-Violation at Category Boundaries 57
    6.2 SUMMARY OF QUALITATIVE INSIGHTS 57
    6.2.1 Appearance Governs Uncanny Valley Depth. 57
    6.2.2 Behavior Partially Mitigates or Exacerbates Experience. 58
    6.2.3 Untrained Animal Behavior Perceived as Threat. 58
    6.2.4 Fur Alone Reduces Metallic Harshness but Retains Machine Cues. 58
    6.2.5 Practical Design Recommendations 58
    6.3 INTEGRATED INTERPRETATION 58
    6.3.1 Consistency and Complementarity Across Methods 58
    6.3.2 Theoretical Contributions and Model Alignment 59
    6.3.3 Reflections on Research Questions and Hypotheses 59
    6.4 LIMITATION 60
    6.4.1 Sample Size and Participant Composition 60
    6.4.2 Restricted Appearance Conditions 60
    6.4.3 Laboratory Environment and Ecological Validity 60
    6.4.4 Reliance on Self-Report Measures 61
    6.5 FUTURE WORK 61
    6.5.1 Broader and More Diverse Sample 61
    6.5.2 Expanded Appearance and Behavioral Conditions 61
    6.5.3 Multimodal Measurement 61
    REFERENCES 63
    APPENDIX A: MODIFIED GODSPEED QUESTIONNAIRE 65
    APPENDIX B: LINKS TO MATERIALS 67
    APPENDIX C: ITEMS IN 2ND PILOT TEST 68

    Aibo. (n.d.). Aibo. Retrieved March 3, 2025, from https://us.aibo.com/
    ANYbotics. (2023). ANYmal: Autonomous inspection solution. https://www.anybotics.com/anymal
    Bartneck, C. (2008, March 11). The Godspeed Questionnaire Series. Christoph Bartneck, Ph.D. https://www.bartneck.de/2008/03/11/the-godspeed-questionnaire-series/
    Bartneck, C., Kulić, D., Croft, E., & Zoghbi, S. (2009). Measurement Instruments for the Anthropomorphism, Animacy, Likeability, Perceived Intelligence, and Perceived Safety of Robots. International Journal of Social Robotics, 1(1), 71–81. https://doi.org/10.1007/s12369-008-0001-3
    Boston Dynamics. (2023). Spot for industrial inspection. https://www.bostondynamics.com/products/spot
    Chan, Y.-C., & Hauser, E. (2023). Understanding Reactions in Human-Robot Encounters with Autonomous Quadruped Robots. Proceedings of the Association for Information Science and Technology, 60(1), 86–97. https://doi.org/10.1002/pra2.771
    Dillow, C. (2012, September 10). DARPA's robotic pack mule LS3 gets field-tested. Popular Science. https://www.popsci.com/technology/article/2012-09/darpas-robotic-pack-mule-ls3-gets-field-tested/
    Franke, T., Attig, C., & Wessel, D. (2019). A personal resource for technology interaction: development and validation of the affinity for technology interaction (ATI) scale. International Journal of Human–Computer Interaction, 35(6), 456-467.
    Friston, K. (2010). The free-energy principle: a unified brain theory?. Nature reviews neuroscience, 11(2), 127-138. https://doi.org/10.1038/nrn2787
    Hashimoto, N., Hagens, E., Zgonnikov, A., & Lupetti, M. L. (2024). Safe Spot: Perceived safety of dominant and submissive appearances of quadruped robots in human-robot interactions (No. arXiv:2403.05400). arXiv. https://doi.org/10.48550/arXiv.2403.05400
    Hauser, E., Chan, Y.-C., Bhalani, R., Kuchimanchi, A., Siddiqui, H., & Hart, J. (2023). Influencing Incidental Human-Robot Encounters: Expressive movement improves pedestrians’ impressions of a quadruped service robot (No. arXiv:2311.04454). arXiv. https://doi.org/10.48550/arXiv.2311.04454
    Ho, C. C., & MacDorman, K. F. (2010). Revisiting the uncanny valley theory: Developing and validating an alternative to the Godspeed indices. Computers in Human Behavior, 26(6), 1508-1518.
    IEEE Spectrum. (2020, June 19). Boston Dynamics’ Spot robot goes on sale. https://spectrum.ieee.org/boston-dynamics-spot-robot-sale
    Joy for All Companion Pet Pup. (n.d.). Ageless Innovation LLC. Retrieved March 3, 2025, from https://joyforall.com/pages/joy-for-all-companion-pet-pup
    Kätsyri, J., Förger, K., Mäkäräinen, M., & Takala, T. (2015). A review of empirical evidence on different uncanny valley hypotheses: support for perceptual mismatch as one road to the valley of eeriness. Frontiers in psychology, 6, 390. https://doi.org/10.3389/fpsyg.2015.00390
    Löffler, D., Dörrenbächer, J., & Hassenzahl, M. (2020). The Uncanny Valley Effect in Zoomorphic Robots: The U-Shaped Relation Between Animal Likeness and Likeability. Proceedings of the 2020 ACM/IEEE International Conference on Human-Robot Interaction, 261–270. https://doi.org/10.1145/3319502.3374788
    MacDorman, K. F., & Ishiguro, H. (2006). The uncanny advantage of using androids in cognitive and social science research. Interaction Studies, 7(3), 297–337. https://doi.org/10.1075/is.7.3.03mac
    Melson, G. F., Kahn, P. H., Beck, A., Friedman, B., Roberts, T., & Garrett, E. (2009). Robots as dogs?–Children’s interactions with the robotic dog AIBO and a live Australian Shepherd. Computers in Human Behavior, 25(2), 303–317. https://doi.org/10.1016/j.chb.2008.10.001
    Mori, M. (1970). The uncanny valley. Energy, 7(4), 33-35.
    Nomura, T., Suzuki, T., Kanda, T., & Kato, K. (2006). Measurement of negative attitudes toward robots. Interaction Studies, 7(3), 437–454.
    PARO Therapeutic Robot | Photo Gallery. (n.d.). Retrieved March 3, 2025, from http://www.parorobots.com/photogallery.asp
    Schwind, V., Leicht, K., Jäger, S., Wolf, K., & Henze, N. (2018). Is there an uncanny valley of virtual animals? A quantitative and qualitative investigation. International Journal of Human-Computer Studies, 111, 49–61. https://doi.org/10.1016/j.ijhcs.2017.11.003
    Unitree Robotics. (2024). Unitree Go2: Next-generation consumer-level quadruped robot. https://www.unitree.com/go2
    Vincent, J. (2021, October 13). This robot dog is now armed — and you can’t do anything about it. The Verge. https://www.theverge.com/2021/10/13/22723548/robot-dog-ghost-robotics-armed-weapon
    Wada, K., Shibata, T., Saito, T., & Tanie, K. (2005). Psychological and social effects of one year robot assisted activity on elderly people at a health service facility for the aged. Proceedings of the IEEE International Conference on Robotics and Automation, 2785–2790. https://doi.org/10.1109/ROBOT.2005.1570535

    無法下載圖示 全文公開日期 2031/02/22
    QR CODE
    :::