Advanced robotic cognition and capabilities for human-robot Interactive tasks in real life settings

先进的机器人认知和能力，用于现实生活中的人机交互任务

• 批准号: 354765-2007
• 负责人: Michaud, François
• 金额: $ 6.89万
• 依托单位: Université de Sherbrooke
• 依托单位国家: 加拿大
• 项目类别: Special Research Opportunity Program - Project
• 财政年份: 2010
• 资助国家: 加拿大
• 起止时间: 2010-01-01 至 2011-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=459290
• 关键词: Advanced; robotic; cognition; capabilities; human

In designing advanced robotic systems, the main challenge is to integrate in a working system all the necessary elements (from mechatronics to sensing, computing, deciding and interacting modalities) required on a mobile robot to fulfill a useful role in real-life settings. Such a design requires a broad set of underlying expertises, usually addressed separately in distinct research laboratories. A unique collaboration opportunity between the Univ. Sherbrooke and Indiana Univ. Bloomington, Arizona State Univ. and Stanford Univ. will make it possible to integrate contributions on robust natural language interaction, affective computing, goal languages, logic and planning and cognitive architecture into an advanced omnidirectional humanoid robotic platform made by the Canadian group. This platform will be capable of changing its height, do gestures and facial expressions, feel the ground through its locomotion components and interact with people and the world with two arms and grippers through force sensing using a new type of actuators. It will use a sound source localization, tracking and separation system as pre-processing for speech recognition and dialogue manage- ment, goal generation and affective evaluation from linguistic and non-linguistic cues. It will also be capable to plan what it must do in highly dynamic and changing conditions of the real world. Computational architectures coming from engineering and cognitive modeling perspectives will be validated using this experimental testbed in realistic operating conditions. The integrated result will be demonstrated at the International RoboCup@Home competition, focusing on real-world applications and human-machine interaction with autonomous robots assisting humans in everyday life. Three Canadian researchers will be funded in this project, training 3 Ph.D., 1 Master's, 3 undergraduate trainee and 1 research professional. This project will produce one of the most advanced autonomous and interactive mobile robot ever built, with groundbreaking results on human-robot interaction and intelligent architectures, making Canada take on a leadership role in new emerging fields of robotics with potential uses in robotic assistive and interactive technologies.

在设计先进的机器人系统，主要的挑战是在一个工作系统中集成所有必要的元素（从机电一体化的传感，计算，决策和交互方式）所需的移动的机器人，以履行在现实生活中的设置一个有用的角色。这样的设计需要广泛的基础专业知识，通常在不同的研究实验室单独解决。舍布鲁克大学和印第安纳州大学、布卢明顿大学、亚利桑那州立大学和斯坦福大学之间的一个独特的合作机会，将使人们有可能将强大的自然语言交互、情感计算、目标语言、逻辑和规划以及认知架构等方面的贡献整合到加拿大集团制造的一个先进的全方位人形机器人平台中。该平台将能够改变其高度，做手势和面部表情，通过其运动组件感受地面，并通过使用新型致动器的力传感，用双臂和抓手与人和世界互动。它将使用声源定位，跟踪和分离系统作为语音识别和对话管理，目标生成和情感评估的语言和非语言线索的预处理。它还将能够计划在真实的世界的高度动态和变化的条件下必须做什么。计算架构来自工程和认知建模的角度将使用这个实验测试平台在现实的操作条件下进行验证。综合结果将在国际RoboCup@Home比赛中展示，重点是现实世界的应用和人机交互，自主机器人在日常生活中帮助人类。该项目将资助三名加拿大研究人员，培养3名博士，硕士1人，本科生3人，研究生1人。该项目将生产有史以来最先进的自主和交互式移动的机器人之一，在人机交互和智能架构方面取得突破性成果，使加拿大在新兴的机器人领域发挥领导作用，在机器人辅助和交互技术方面具有潜在的用途。