本計畫建立智慧型生命徵象辨識與救援排程之即時遠端監控系統的設計與實現，應用於救援機器人於災難現場之遠端監控。 第一年本計畫將建立纜線對接通訊系統設計與切換機制、生命徵象、環境狀態指標分析與判讀、圖形概念設計與建立、硬體裝置整合。通訊採用纜線傳輸以免受到災難現場環境影響；考慮受難者與機器人所處的環境狀態影響將生命徵象判讀分為六種等級；操作介面採用圖形化概念讓救援人員在監控端操作更易熟練；設計具高機動性箱型硬體架構與設備擺放空間配置。 第二年本計畫將完成無線行動隨意網路設計、救援策略之半自主控制模式、救援軟體平台實現模擬、模組化驗證平台建立。透過802.1轉接器連接無線通訊；整合各子計畫指標擬定救援策略與機器人半自主模式；指標資料顯示於救援軟體平台，提供受難者目前狀況與位置；組裝硬體設備於箱子內部建立操控平台。 第三年本計畫將完成混合式通訊系統整合、智慧型半/全自主救援排程模組整合、介面最佳化整合、裝置完整性與安全性驗證。驗證通訊系統切換機制確保資料傳輸的穩定與正確性；通訊受限下機器人須自行判斷受難者狀態；整合所有救援人員需要使用的資訊項目於介面平台顯示；將遠端介面平台及機器人操控介面與箱型裝置整合。 We propose a three year project to design and implement a real time remote monitoring and control system in a portable suitcase for vital sign recognition and rescue task scheduling in rescue robot applications. The focus of this subproject is four folds: i) data communication; ii) task scheduling; iii) human computer interface; and iv) hardware integration. In the first year (2011.8~2012.7), we propose a tethered docking communication system to which is advantageous three fold: i) ensure the communication bandwidth for feedback of video signals and output of control signals to the remote robot; ii) provide power to the robot and extend time of task execution without needing to replace batteries; and iii) act as a life-line to retrieve the robot and/or survivors. The docking system will detach when the tether is constraint and switch to wireless communication mode. Meanwhile, the docking system can reattach for power charge, etc. We design a vital sign prioritizing algorithm to ensure that the maximum number of victims are recovered with minimum search time in the search and rescue task. The workflow for GUI proposed to ensure to minimize the learning time of the monitoring and control system and maximize the efficiency of data representation without confusion of data interpretation. We propose integration of two multi-touch industrial com-puters along with multi-directional joysticks in a disaster environment robust compact suit-case for remote monitoring and semi-autonomous operation of the remote robot. In the second year (2012.8~2013.7), we design and implement a wireless mobile ad hoc network (MANET) based on 802.11g energy efficiency of transmission power control to achieve minimal power consumption of the wireless system. For rescue task scheduling, we integrate environmental indices with vital sign indices to ensure optimal search and rescue scheduling. For human computer interface, we propose a markerless augmented reality (AR) approach to display objects which are not perceivable by camera based images. Finally in the third year (2013.8~2014.7), we verify the hybrid communication scheme where tethered is used for global long-distance communication with wireless as local short-distance communications to cope with various environmental conditions. In addition, a quality of service (QoS) criteria based hopping scheme is introduced for network topology of the MANET. For rescue task management, we implement a full autonomous search and rescue task to validate the results of the previous two years. In addition, we optimize the human computer interface to make sure efficiency of manual operation is optimized. Finally, all technologies this project presents will be developed from a modular point of view to provide the rescue robot ability to extend the search and rescue task to multi-robot cooperation.
