Advanced tools for flexible robotics

该项目旨在开发智能解决方案，通过使用多功能夹具的机器人握住的工具或部件，实现有效、灵活和自适应的任务性能，从而引领新一代机器人系统的发展。开发的解决方案将包括正确抓取或重新抓取物体或工具的能力，确定给定抓取任务的可行性，根据需要规划重新抓取轨迹，适应环境变化，并有效执行可能需要位置，力量或混合控制能力的任务。为此，将开发和制造两种新的机器人手概念。其中一只手将被设计用来操作常见的工业物品或工具，而另一只手将基于拟人化的方法。具有扩展功能的多模态触觉传感组件也将被开发并集成到机器人手中。这些传感器将能够检测抓取的物体，与环境接触和其他刺激引起的振动。将开发信号处理技术，以便将传感器输出转换为可用于抓取和任务执行场景的高级信息。例如，机器人将能够根据抓取工具的姿态确定是否可以使用工具执行给定任务，并在需要时计划重新抓取动作，从而大大提高机器人机械手的灵活性，适应性和自主性。此外，将开发一种力/扭矩传感器，以便提供确定外部负载并相应地指导任务的有效手段。最后，所有结果将集成到功能原型中，用于产生令人信服的行业相关演示所开发的解决方案。更灵活和适应性强的机器人有可能通过开辟新的可能性来改写制造业自动化的规则，例如全自动小批量生产、自主敏捷制造、可靠的自动质量控制等等。加拿大制造业非常需要这些能力。

This project aims to develop intelligent solutions for effective, flexible and adaptive task performance with tools or parts held by a robot using a versatile gripper, thereby leading the way to a new generation of robotic systems. The solutions developed will include the capability to properly grasp or regrasp objects or tools, to determine the feasibility of a given task with a given grasp, to plan regrasping trajectories if required, to adapt to changes in the environment and to effectively perform tasks that may require position, force or hybrid control capabilities. To this end, two new concepts of robotic hands will be developed and built. One of the hands will be designed to manipulate common industrial objects or tools while the other will be based on an anthropomorphic approach. Multi-modal tactile sensing components with extended capabilities will also be developed and integrated into the robotic hands. These sensors will be capable of detecting objects grasped, vibrations induced by contacts with the environment and other stimuli. Signal processing techniques will be developed in order to transform the sensor outputs into high-level information that can be used in a grasping and task performance scenario. For example, the robot will be capable to determine whether a given task can be performed with a tool based on the pose of the grasped tool and to plan regrasping maneuvers if required, thereby greatly improving the flexibility, adaptability and autonomy of robotic manipulators. Also, a force/torque sensor will be developed in order to provide an effective means of determining the external loads and guide the tasks accordingly. Finally, all results will be integrated into functional prototypes that will be used to produce convincing industry-relevant demonstrations of the solutions developed. More flexible and adaptable robots have the potential to rewrite the rules of manufacturing automation by opening new possibilities such as fully automated small series, autonomous agile manufacturing, reliable automatic quality control, and many others. These capabilities are in great need in the Canadian manufacturing industry.