- 學位論文
機器人動作語意與情感訊息之探討
A Study on Semantic and Emotional Messages in Robot Movements
摘要
隨著機器人開發技術的日趨完善與普及應用，機器人也開始進入人類的現實生活當中，使用者與機器人的互動過程中，機器人如何透過肢體動作來傳達訊息，並且與使用者產生互動，而機器人與其使用者之間的人機互動能否跟人際間互動一般，藉由動作來傳遞HRI訊息或是表現情緒反應?以及使用者是否能了解機器人動作所代表的語意？本研究目的為：(1) 蒐集現有的機器人動作指令，並分析機器人動作的構成要素。(2) 解析動作中影響使用者對該動作的語意判斷與情感認知的要素為何。(3) 透過動作構成要素與影響使用者語意或感受的關連性研究，作為未來以動作來傳達機器人訊息之HRI介面設計之參考。 本研究著重於機器人與人互動之動作及行為表現，主要利用拉邦動作分析(Laban)理論與感性工學為基礎，透過動作相關的形容詞語彙搜集、群聚分析、感性實驗設計、並將實驗結果數量化I類與因素分析等，來解析機器人動作構成要素與使用者感受到的語意或感性訊息之關連性，最後設計出3組機器人動作規則並將結合情境，進行SUS系統使用性尺度量表評估其介面使用性。 本研究最後結論發現，藉由數量化I類分析，拉邦動作分析中的「勁」與情緒之間關聯性。其次是從機器人傳達訊息中發現共有3個因素，分別為「主動」、「遲疑」、「不積極」。在動作與傳達訊息模式中得知，使用者最能接受的動作規則是「挑選訊息與動作之間最高數值為主要的代表性動作」，它是大多數人認同此動作比較符合此訊息的代表性，也不用重新學習動作與訊息之間的關連性。
並列摘要
In recent years, advances in robotic technologies have motivated researchers calling for better human robot interaction (HRI) to engage users in a more socially meaningful way. Research questions, such as how to use robots’ movements to convey of messages or to express emotional response for better communication in HRI and how users interpret robots’ movements have become important issues in design research and practice. Hence, the purpose of this study consists: (1) Collecting the existing robot motion commands, and analyzing the component of the robot movement. (2) Analyzing the factors that influence users understanding toward actions. (3) By the study of connections between acting organization and users’ understanding, to be the reference materials of robot HRI interface design that can correctly deliver message. The study is focusing on the interaction between robot and human, based on the LMA theory and perceptual learning, collecting by the adjective that related to action, through cluster analysis, perceptual experimental design, statistical analysis, and made the result of experiment Qualification I and Factor analysis, etc., to explain the connection between robot action organization elements and the user understanding. In the end, we design three set of robot movement rules to integrate with the scenarios, and by using the SUS(system usability scale) to estimate the use value of interface. In conclusion, the study found out that the relation between “Effort” and emotion in the Laban Movement Analysis could be analyzed by Qualification I. Second, there are three factors while robot delivering the messages, which are "active", "delay," and "positive." In the mode of movements and message delivery, the most acceptable movement rule by users is "to choose the highest number between the message and the action as the main representative action," which most people agree that this action is more in line with the message, and no need to learn the relationship between the actions and messages.