Automated welding: 3D reconstruction of workspace for safe robot navigation

Partner:

IKNOWHOW

Problem:

IKNOWHOW, a provider of custom-built automation solutions, used to make a 3D reconstruction of the workspace by taking photos with a DSLR camera from different angles. These photos were then “stitched together” to reconstruct the environment using the photogrammetry method. However, this method contained errors and was highly inefficient as it took about 20 minutes, which severely affected the quality of the entire workflow.

Solution:

PhoXi 3D Scanner M

Because of the above limitation, IKNOWHOW decided to use a UR10e manipulator equipped with a welding torch, EVT’s Saturn laser seam tracker, a PhoXi Scanner model M from Photoneo, an HTC Vive 6DOF motion tracker, and an AI module for seam recognition implementing UMNAI’s XAI. 

The PhoXi 3D Scanner creates a 3D representation of the workspace, which is used for obstacle-free path planning and safe robot navigation around obstacles and welding pieces. Using the Vive tracker, the user indicates where the welding seam lies. Then the robotic arm can accurately track the seam using the tracker and weld alongside.

Certified System Integrator Program

Set Yourself at the Forefront of the Global Vision Market

Vision system integrators certified by A3 are acknowledged globally throughout the industry as an elite group of accomplished, highly skilled and trusted professionals. You’ll be able to leverage your certification to enhance your competitiveness and expand your opportunities.

GET CERTIFIED

Discussion:

IKNOWHOW chose the Photoneo PhoXi 3D Scanner for this project based on the high quality of 3D point clouds the device provides. Other factors that played a role in the decision were the size of the workspace and the scanner’s scanning volume, as well as easy calibration and installation.

The results were impressive. The main advantages of this new method include:

• Time savings – the previous photogrammetry method took about 20 minutes, now it is a matter of seconds
• High accuracy and resolution
• The produced point cloud is only ~1 millimeter off in size and placement compared to the real physical piece, which is very satisfactory
• Easy calibration