Development of Self-Adaptable Visual-Servo Pick-and-Place Robot Technology for Manufacturing Automation

制造自动化自适应视觉伺服取放机器人技术的开发

• 批准号: 543612-2019
• 负责人: Zhu, ZhengHong
• 金额: $ 1.82万
• 依托单位: York University
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2019
• 资助国家: 加拿大
• 起止时间: 2019-01-01 至 2020-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=680514
• 关键词: Development; Self; Adaptable; Visual; Servo

Pick-and-place Robots have been widely used to automate laborious and repetitive tasks by human operators to increase productivity, lower labour costs and reduce workplace injury claims, while ensuring consistent quality of products at high speed. However, specially designed pick-and-place robots for certain products and processes can no longer fit the current fast changing manufacturing process and products, where different products may be handled. As a result, many shops of small and medium auto parts manufacturers still retain human operators at the end of automated assembly lines to pick up and sort finished parts. The objective of the proposed research is to develop a programming framework for self-adjustable visual-servo pick-and-place robotic technology that can be readily implemented into commercial industrial robots. The research will integrate the computer vision and robotic control technologies with artificial intelligence and machine learning to replace human operators. The computer vision will be programmed with artificial intelligence tools to enhance the ability in successful identification and recognition of parts' shape and pose in shop environment. Then, a path for the end-effector of the robot will be planned automatically for a pick-and-place task, subject to the constraint of space and operation requirements. The visual and path information will be fed dynamically into the inner robot control system to execute the task. The outcome of the proposed research provides a last-mile delivery of robotic technologies to small and medium sized auto parts manufacturers, which is currently not covered by major manufacturers of industrial robots. TiGold Automotive North America Co Ltd in Ontario has identified this niche market and supports the proposed development. The proposed technology will increase the productivity and competitability of Ontario's automotive industry sector. It will not only increase jobs to our industrial partner by adding more service capability but also the local suppliers. In addition to the economic and technical benefits to Canada outlined above, this project will train three HQP, benefit the industrial partner, its employees, local suppliers and service providers.

取放机器人已广泛用于自动执行人类操作员的繁重和重复性任务，以提高生产率、降低劳动力成本并减少工伤索赔，同时确保高速生产的产品质量始终如一。然而，针对某些产品和工艺专门设计的取放机器人已无法适应当前快速变化的制造工艺和产品，其中可能处理不同的产品。因此，许多中小型汽车零部件制造商的车间仍然在自动化装配线末端保留人工操作员来拾取和分类成品零件。拟议研究的目的是开发一种用于自调节视觉伺服拾放机器人技术的编程框架，该框架可以很容易地应用于商业工业机器人。该研究将计算机视觉和机器人控制技术与人工智能和机器学习相结合，以取代人类操作员。计算机视觉将采用人工智能工具进行编程，以增强在车间环境中成功识别和识别零件形状和姿势的能力。然后，根据空间和操作要求的限制，将自动规划机器人末端执行器的路径，以执行拾放任务。视觉和路径信息将动态地输入内部机器人控制系统以执行任务。拟议研究的成果为中小型汽车零部件制造商提供了机器人技术的最后一英里交付，目前主要工业机器人制造商尚未涵盖这一点。位于安大略省的 TiGold Automotive North America Co Ltd 已确定了这一利基市场并支持拟议的开发。拟议的技术将提高安大略省汽车工业部门的生产力和竞争力。它不仅会通过增加更多的服务能力来增加我们的工业合作伙伴的就业机会，而且还会增加当地供应商的就业机会。除了上述为加拿大带来的经济和技术利益外，该项目还将培训三名总部人员，使工业合作伙伴、其员工、当地供应商和服务提供商受益。