机器人引导：自动车轮安装

在设计可以提高性能、流程和质量的应用程序时，需要解决许多挑战。这些挑战涉及转子和安装轮胎的过程：

• 超过 60 种不同的边缘用于不同类型的表面（深色、哑光、光泽），由于照明解决方案对不同类型表面的影响，这使得 2D 相机系统难以捕捉。
• 螺栓预装在每个路由器上。每个螺栓帽的表面积很小，这意味着需要一个 3D 高分辨率视觉系统来准确定位每个螺栓帽的点云。
• 在安装过程中，转子随机旋转，这意味着每次安装时螺栓都处于不同的位置，因此需要一种能够识别螺栓位置的解决方案。
• 该解决方案还需要能够进行 3D 匹配或大型点云，因为转子可以与车辆一起在两个方向上旋转 15 度。
• 除了这些技术因素外，该部件很重，并且整个点云抓取和处理的周期时间有限，只有 3.5 秒。

Visio NERF 实施的解决方案使用高分辨率蓝光 3D 相机。 3D 传感器组合成一个蓝色 3D Vision Robot 来指导解决方案软件。机器人技术中使用了两个带有七轴线性滑轨的 FANUC 机器人。使用该系统，车轮和转子的点云可以非常准确地与相应的 CAD 模型相匹配。

• [Efficiency Improvement - Productivity]
  

  3D vision systems are capable of capturing high-quality point clouds of different types of materials. The system is robust to colour and brightness parts: tires with a dull surface, rotors with a bright glossy surface. In addition, the design of the sensor makes it possible to simultaneously acquire parts made of multiple materials in the same scene. The visual processing cycle is 3.5 seconds, the efficiency is greatly improved, the initial problem has been solved, and the whole process no longer requires any worker participation.

• Depending on the manufacturer's production needs, the complete solution can scale from a few hundred to thousands of vehicles per day. The customer's current production is 250 vehicles per day and the uptime of the wheel mounting system is consistently 99.5%. The program picks up the wheel and mounts it to the rotor, with multiple functions, including those needed to handle different surfaces, mounting locations, and bolts.