arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.14388cs.RO
收藏

通过操作员技能识别、建模与人机协作改进焊接机器人化:实验方案实施

Improving Welding Robotization via Operator Skill Identification, Modeling, and Human-Machine Collaboration: Experimental Protocol Implementation

Nantes Université(南特大学);École Centrale Nantes(南特中央理工学院);CNRS(法国国家科学研究中心);CETIM(法国技术与工业设备中心);Université Rennes 2(雷恩第二大学)

Antoine Lénat, Olivier Cheminat, Damien Chablat, Camilo Charron

AI总结:

本文面向工业5.0中仍难机器人化的手工制造任务,以焊接为案例,提出并实施一种结合操作员技能识别、数字化建模与人机协作的实验方案,旨在形成可迁移至其他领域的方法论。

摘要:

未来工业,也称为工业5.0,旨在实现生产工具现代化、车间数字化,并培育企业内部宝贵的人力资本。工业5.0的实现离不开培养一支不仅技术娴熟而且技能和知识得到增强的劳动力队伍。具体而言,协作机器人技术在自动化繁重或重复性任务方面发挥着关键作用,使人类的认知功能能够为质量和创新做出贡献。然而,在手工制造中,其中一些任务在不牺牲质量的情况下仍然难以自动化。在某些情况下,这些任务要求操作员动态组织其心理、感知和手势活动。换句话说,这些技能尚未得到充分解释和数字化建模,无法让工业环境中的机器即使以近似方式再现它们。焊接中的某些任务就是一个完美的例子。借鉴认知与发展心理学、职业教学法和协作机器人研究的成果,我们的工作旨在找到一种对手工制造技能进行数字化建模的方法,以增强那些仍然难以机器人化的任务的自动化程度。以焊接为例,我们力求开发、测试并部署一种可迁移到其他领域的方法论。本文的目的是介绍用于实现这些目标的实验设置。

英文摘要:

The industry of the future, also known as Industry 5.0, aims to modernize production tools, digitize workshops, and cultivate the invaluable human capital within the company. Industry 5.0 can't be done without fostering a workforce that is not only technologically adept but also has enhanced skills and knowledge. Specifically, collaborative robotics plays a key role in automating strenuous or repetitive tasks, enabling human cognitive functions to contribute to quality and innovation. In manual manufacturing, however, some of these tasks remain challenging to automate without sacrificing quality. In certain situations, these tasks require operators to dynamically organize their mental, perceptual, and gestural activities. In other words, skills that are not yet adequately explained and digitally modeled to allow a machine in an industrial context to reproduce them, even in an approximate manner. Some tasks in welding serve as a perfect example. Drawing from the knowledge of cognitive and developmental psychology, professional didactics, and collaborative robotics research, our work aims to find a way to digitally model manual manufacturing skills to enhance the automation of tasks that are still challenging to robotize. Using welding as an example, we seek to develop, test, and deploy a methodology transferable to other domains. The purpose of this article is to present the experimental setup used to achieve these objectives.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑