具老手便捷性之車載語音人機介面設計

Human-Machine Interface Design of In-Vehicle Speech Interaction System with Expert Convenience

古佩盈

指導教授 ： 黃雪玲

國立清華大學/工學院/工業工程與工程管理學系/碩士(2013年)

摘要

目前由工研院所研發的車用語音系統(Talking Car)的設計較市面上車用語音系統擁有更佳的安全性，但其引導性與便捷性之部分尚未被探討。本研究主要是設計具有老手便捷性之語音互動介面，並規劃出一套符合行車安全法規的「全語音系統」，透過實驗進行可用度評估，蒐集和歸納受測者對系統的行為與反應數據，同時觀察受測者在使用語音介面時，行車時注意力分散的情況以確保Talking Car對於行車之安全性不會造成負面影響。
本研究為了發展具有老手便捷性之語音互動介面，前後共歷經三個階段的駕駛模擬實驗，共16名受試者；受試者需在開車的狀況下執行我們所給予的Talking Car操作任務，任務內容選擇以車用語音系統中最常被使用的音樂與廣播功能為主。本研究之實驗包含兩項自變數－使用者對於Talking Car之熟悉程度與Talking Car之模式，並區分為新手、進階新手及老手三個等級。實驗中會利用攝影機拍攝受測者的操作狀況，記錄受測者操作任務時前方出現”STOP!!”剎車訊息時踩剎車的反應時間、和完成任務時間，並在實驗後請受測者完成語句煩躁程度評分、心智負荷問卷和主觀問卷，以評量使用者對系統語句的煩躁程度、工作負荷和主觀感受。
實驗結果發現，新手模式的系統語句經過兩階段簡化後成為老手模式，根據實驗研究評估，老手模式系統語句煩躁程度明顯較新手模式低；同時也發現使用者中斷語句之位置可作為系統自動切換至老手模式之依據。我們以緊急煞車時的反應時間作為衡量指標，評估Talking Car系統對於行車安全性的影響，發現受試者的反應時間皆在國際上普遍認定的安全範圍，亦即2.5秒以內，因此Talking Car並不會成為駕駛行車時的威脅。

關鍵字

老手便捷性 ； 車用語音介面 ； 可用度評估 ； 可用度評估

並列摘要

This research aimed to investigate and explore expert user interface design principle in adaptive user interface of in-vehicle full voiced-based interface. The voice-based interaction interface with expert convenience was developed step by step through three stages of experiments. The voice-based interface called Talking Car novice interface has been designed before. By conducting those driving simulation experiments, number of characters in Chinese the expert could tolerate of Talking Car expert interface would be found and ensure there are sufficient users’ operation efficiency and traffic safety.
There were sixteen participants who had car driving licenses and were novice user of in-vehicle full voiced-based interface. They would be trained into expert users by going through the experiments. At first, the novice participants used Talking Car novice interface to carry out the experiment, then the number of characters the advanced-novice users could tolerate of Talking Car would be found. According to the result, the length of speech by Talking Car would be adjusted to fulfill advanced-novice users’ requirements, and then it change into Talking Car advanced-novice interface. Then, the Talking Car expert interface would be developed after second stage experiment (advanced-novice user using advanced-novice interface). After that, those participants would carry out the third stage experiment to verify the usability as well as the implications on operation efficiency and traffic safety by using Talking Car expert interface. In driving simulated experiments, a camera was set to capture the subject’s operations, reaction time of the brakes when “STOP!!” message showed on the display and the task completion time. After the experiment finished, subjects would be asked to complete three questionnaires, included NASA-TLX, subjective and impatient-degree questionnaire, to evaluation subjects’ workload, satisfaction and the degree of impatient of system speeches.
After three stage experiments, the results suggested the suitable sentence length of each interface. According to the experiment result, we could also define that novice interface sentence would be changed into advanced speech sentence when it has been interrupted for 6 times, and advanced-novice interface would be changed into expert interface sentence when it has been interrupted for 5 times. In addition, the sentence of expert interface resulted in lowest degree of impatient, required less operation time and mental workload than that of other interfaces. Therefore, drivers not only feel more convenience but also have a safer driving condition when they are controlling the Talking Car Expert Mode.

並列關鍵字

expert user interface ； adaptive user interface ； in-vehicle full voice-based interface ； Talking Car

參考文獻

Chen, F., Jonsson, I.-M., Villing, J. and Larsson, S. ( 2010). Applications of Speech Technology in Vehicles. In Fang Chen and Kriistina Jokinen, (eds.): New Trends in Speech Based Interactive Systems (pp. 195-219). New York: Springer.

American Association of State Highway and Transportation Officials. A Policy on Gecwnemic Design of Highnwysand Streets. Washington, D.C.:AASHTO, 1984

Angkititrakul, P., Petracca, M., Sathyanarayana, A., Hansen, J.H.L. (2007). UTDrive: driver behavior and speech interactive systems for in-vehicle environments. IEEE Intelligent Vehicles Symposium, 566–569.

Benyon, D. and Murray, D. (1993). Applying user modeling to human-computer in-teraction design. Artificial Intelligence Review. 7, 199-225.

Chang JC, Lien A, Lathrop B, Hees H., (2009). Usability evaluation of a Volkswagen Group in-vehicle speech system. Proceedings of the 1st international conference on automotive user interfaces and interactive vehicular applications (Automo-tiveUI’09), 137–144. doi:10.1145/1620509.1620535