MOST - Task-relevant perception and control for human-oriented operation of mobile manipulators in semi-structured environments

MOST - 半结构化环境中移动机械手人性化操作的任务相关感知和控制

• 批准号: 506987-2017
• 负责人: Jeon, Soo
• 金额: $ 13.47万
• 依托单位: University of Waterloo
• 依托单位国家: 加拿大
• 项目类别: Strategic Projects - Group
• 财政年份: 2017
• 资助国家: 加拿大
• 起止时间: 2017-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=642037
• 关键词: MOST; Task; relevant; perception; control

The rise of e-commerce has made logistics (material handling) the fastest-growing industry today with the annual growth rate of 10-15% worldwide. The digitalization of consumer markets is also calling for new paradigm of manufacturing to cope with increasing demands for short-life cycle products and burst manufacturing. The immediate consequence of this trend is the urgent need for a faster and more efficient operation of large scale material handling plants such as sorting facilities for parcels, distribution centers for digital retail and machine tool factories for discrete-part manufacture. These indoor areas are "semi-structured" in the sense that fixed installations (shelves, workstations) are blended with dynamic unpredictable entities (human workers, indoor vehicles) to transfer numerous objects in different sizes, shapes and properties. This research brings together experts in control, sensing, HRI (Human-Robot Interaction) and machine intelligence from Canada and Taiwan to develop intelligent mobile manipulators that can operate side-by-side with humans in semi-structured environments. In order for robots to collaborate with humans in an intuitive and reliable manner, we emphasize the formal methods for HRI and shared autonomy. New sensing and decision making algorithms will be developed to allow robots to navigate through dynamic environments safely and manipulate objects with human-like dexterity. This research builds on existing infrastructure including the mobile manipulator testbeds in both research groups as well as the new RoboHub facility at Waterloo, a high-tech indoor space for robot testing and validation. The demand for more labor is increasing in logistics and material handling, yet the available workforce will decrease due to shrinking population and aging. This research will lay the foundation for next generation of material handling facilities where mobile manipulators can coexist and operate seamlessly with human workers. The new knowledge and techniques from this research will also foster progresses in other areas including perception, autonomous navigation, biomechatronics, personal and healthcare robotics. This project is submitted under the international agreement between NSERC and MOST.

电子商务的兴起使物流（物料搬运）成为当今增长最快的行业，全球年增长率为10-15%。消费市场的数字化也呼唤新的制造模式，以科普对短生命周期产品和突发制造的日益增长的需求。这一趋势的直接后果是迫切需要更快，更有效地运营大型物料处理工厂，如包裹分拣设施，数字零售配送中心和离散零件制造机床工厂。这些室内区域是“半结构化”的，在这个意义上，固定的设施（货架，工作站）与动态不可预测的实体（人类工人，室内车辆）混合在一起，以转移不同大小，形状和属性的许多物体。这项研究汇集了来自加拿大和台湾的控制，传感，HRI（人机交互）和机器智能方面的专家，以开发智能移动的机械手，可以在半结构化环境中与人类并肩作战。为了让机器人以直观和可靠的方式与人类合作，我们强调HRI和共享自治的形式化方法。将开发新的传感和决策算法，使机器人能够安全地在动态环境中导航，并以类似人类的灵活性操纵物体。这项研究建立在现有的基础设施上，包括两个研究小组的移动的机械手测试台以及滑铁卢的新RoboHub设施，这是一个用于机器人测试和验证的高科技室内空间。物流和材料处理对更多劳动力的需求正在增加，但由于人口减少和老龄化，可用劳动力将减少。这项研究将为下一代物料搬运设施奠定基础，在下一代物料搬运设施中，移动的机械手可以与人类工人共存并无缝操作。这项研究的新知识和技术也将促进其他领域的进步，包括感知，自主导航，生物机电一体化，个人和医疗机器人。该项目是根据国家经济和环境研究中心与科技部之间的国际协议提交的。