一般大型公共場所或會場的服務人員，常見的服務是導引訪客到想去之目的地，而除了服務人員的協助外，也有單位提供電腦輔助查詢的方式（如政府或醫療單位）以降低人力成本。雖然電腦可以提供輔助導引的功能，但所提供的資訊均為固定，不會隨情境變換而更新，例如：此單位未到開放時間等。
目前智慧型導引機器人的相關研究中，雖然機器人己經可以導引訪客至目的地，卻無法提供訪客該目的地之即時情境資訊。如果導引機器人能具備情境感知的功能，便可提供目的地相關之資訊來協助訪者判斷是否要前往，或者可讓機器人在導引時依據當時的情境即時改變行為。
因此，本論文製作一具有情境感知之智慧型語音導引機器人，使用非特定語者的語音辨識與訪客對話互動，並告知來訪者目的地情境之即時資訊，讓來訪者決定是否要導引至目的地，減輕來訪者的搜尋成本；除此之外，亦系統能讓使用者易於置換地圖和情境，以滿足使用者不同之需求。
實驗結果顯示，該具辭書式情境感知之語音導引機器人在非特定語者之辨識率達到73.37%且可依據情境的變換提供即時資訊，有效的引導使用者至目的地。

In a public place or a conference, a common service is to guide visitors to their destinations accurately. Except the assistances offered by information officers, some places, for example government or medical units, may provide kiosks to reduce the manpower cost for guiding. Although kiosks can provide guiding information, the information is usually fixed and cannot be renewed immediately according to the change of places’ context, for example the destination is not open yet.
At present, researches about guiding robots usually result in robots that can guide visitors to their destinations but cannot provide them with the immediate context-aware information on the course to their destinations. If a guiding robot is context-aware, it can assist a visitor in judging whether to go to the destination or not by using the context given by the robot and can change its own behaviors immediately accordingly.
This thesis proposes a guiding robot system with environment context aware and human interacting capabilities. Environment context is represented by an ontology that can be changed easily for the robot to work in different places. And to interact with different people, the guiding robot is equipped with a speaker-independent voice recognition system that interacts with causal visitors through dialogue, tells the visitors vital information on the course to their destinations, and lets the visitors decide whether to go or not.
Experimental results reveal that the recognition rate of the speaker-independent voice recognition system can reach up to 73.37% correct, and the guiding robot can provide immediate context-aware information. It also shows that the system is effective in guiding visitors to their destination.

[1] Actroid接待機器人, http://en.wikipedia.org/wiki/Actroid, last accessed June 2009
[2] 2005年日本愛知博覽會「接待機器人」, http://www.expo2005.or.jp/tcn/robot/robot_project_01.html, last accessed June 2009
[3] 日本人的機器人情結 http://news.xinhuanet.com/tech/2006-09/29/content_5152254.htm, last accessed June 2009
[4] 智慧型導覽機器人, http://www.robotworld.org.tw/index.htm?pid=12&News_ID=46, last accessed June 2009
[5] 日本愛知博覽會, http://www.expo2005.or.jp/tcn/robot/, last accessed June 2009
[6] 機器人情報網, http://www.robotworld.org.tw/index.htm, last accessed June 2009
[7] NTU-1 智慧型校園導覽互動機器人, http://www.robotworld.org.tw/index.htm?pid=12&News_ID=635, last accessed June 2009
[8] 取樣定理, http://zh.wikipedia.org/w/index.php?title=%E9%87%87%E6%A0%B7%E5%AE%9A%E7%90%86&variant=zh-tw, last accessed June 2009
[9] 王小川, 語音訊號處理(修訂版), 全華科技圖書有限公司,2007
[10] 王漢蘭, 隱藏式馬可夫模型應用於語音訊號辨識之研究, 中原大學資工所碩士論文, 1999
[11] 李佳慧, 不特定語者國語語音字詞辨識系統研究, 中原大學資工所碩士論文, 2001
[12] 李健興, 適用於家庭應用之推論引擎技術研究, 工研院期中報告, 2006
[13] 林文景, 克服樓層障礙之智慧型機械人, 國立成功大學工科所碩士論文, 2006
[14] 林維琦, 古今中外人名語音辨識系統之設計研究, 國立中山大學電機所碩士論文, 2006
[15] 徐世昌, 郵件機器人之路徑規劃與實現, 國立成功大學工科所碩士論文, 2007
[16] 黃鼎堯, 半人形機器人之語音對話導覽系統設計與實現, 國立成功大學電機所碩士論文, 2008
[17] 楊永泰, 隱藏式馬可夫模型應用於中文語音辨識之研究, 中原大學資工所碩士論文, 2000
[18] 楊鎮光, Visual Basic與語音辨識─讓電腦聽話, 松崗電腦圖書有限公司,2002
[19] 鄭吉峰, 台灣地區中文地址語音辨識系統之設計研究, 國立中山大學電機所碩士論文, 2005
[20] 謝銘原, 智慧型接待服務機器人互動功能之研製, 南台科技大學電機所碩士論文, 2005
[21] Ausu eeePC, http://eeepc.asus.com/tw/, last accessed June 2009
[22] Cross-Platform Rich Client Development Framework, http://trolltech.com/products/qt, last accessed June 2009
[23] Cross-validation, http://en.wikipedia.org/wiki/Cross-validation_(statistics)#K-fold_cross-validation, last accessed June 2009
[24] A. Zelinsky, R.A.J., J.C. Byrne and S. Yuta,in Planning Paths of Complete Coverage of an Unstructured Environment by a Mobile Robot. November 2-5, 1993, International Conference of Advanced Robotics - ICAR '93
[25] Frahling, G. and C. Sohler,A fast k-means implementation using coresets, in Proceedings of the twenty-second annual symposium on Computational geometry. 2006, ACM: Sedona, Arizona, USA
[26] Henry Lieberman, T.S., Out of Context, A Curse on Computer Systems That Adapt To, and Learn From,Context,
[27] Huang, X., Y. Ariki, and M. Jack, Hidden Markov Models for Speech Recognition. Edinburgh, Informtion Technology, Series Edinburgh University Press, No.7,1990
[28] Natalya F. Noy , D.L.M., Ontology Development 101: A Guide to Creating Your First Ontology, S.K.S.L.T. Report and S.M.I.T. Report, March 2001
[29] Rabiner, L.R., A tutorial on hidden Markov models and selected applications in speech recognition, in Readings in speech recognition. 1990, Morgan Kaufmann Publishers Inc. p. 267-296
[30] Sayayama, S. and S. Takahashi. On the use of scalar quantization for fast HMM computation. in Acoustics, Speech, and Signal Processing, 1995. ICASSP-95., 1995 International Conference on. 1995
[31] V. R. Algazi, K.L.B., M. J. Ready, D. H. Irvine, C. L. Cadwell and Sang Chung, Transform Representation of the Spectra of Acoustic Speech Segment with Applications－I: General Approach and Application to Speech Recognition, vol.1, No.2, April 1993
[32] Zelinsky, A., Using path transforms to guide the search for findpath in 2D, IJRR 13, Volume 13, Number 4: p. P.315-325, August 1994
[33] Zenlinsky A. and Yuya, S.v., pp.84-93,1993, Reactive planning for mobile robots using numeric potential fieleds, vol.1: p. p.84-93, 1993