Major automotive manufacturer retrofits vision system for productivity gains

Summary

The powertrain plant of a major automotive manufacturer’s powertrain plant was experiencing problems with an existing vision system. By retrofitting the vision system using Apera AI’s Vue robotic vision software, productivity increased, downtime was eliminated, and the precision of the process was improved. This case study was first presented at Automate 2023, hosted by the Association for Advancing Automation (A3).

Front view of the Fanuc / Apera AI robotic cell, showing picking of a transmission case from a cart.

Project scope

This is an existing robotic cell using two industrial robots and a vision system for robotic guidance. The cell is used to pick transmission cases off of a cart and position them in a laser etching station.

In about 50% of picks, the existing vision system was missing picks. The cell was experiencing significant downtime caused by these missed picks. These issues were attributed to the complex geometry of the part, its reflective finish and variable positioning of the cases on the cart.

The goal of the project was eliminate missed picks and cell downtime. Improved positioning of the transmission cases at the laser etching station was a secondary goal.

A major automotive manufacturing used Apera AI's Vue vision software to guide a Fanuc robot in picking and orienting transmission cases.
The complex geometry of the transmission cases is a vision challenge. Apera AI trained the part into its Vue robotic guidance software to create a reliable picking solution.

Project results

The chosen path was to retrofit the existing vision system with Apera AI’s Vue robotic vision software, implemented with three pairs of 2D cameras used with Apera vision systems positioned high above the cell.

Since commissioning, there have been no missed picks. Despite the complex geometry and shiny finish on the part, the parts are better positioned for laser etching. The project team from the customer has said that the project has created considerable annual savings through eliminated downtime and improved robotic cell productivity.

The main area that required adjustment was how the cameras were positioned above the robot to gain a wide field of view. Three pairs of cameras coupled to the same computer running Apera’s Vue software run the system. The project team recommends to other users to consider camera positioning and to check the software of the robot for programming purposes prior to doing your own vision retrofits.

Visualization of the transmission case within Apera's software.
Visualization of the transmission case within Apera AI’s software. By teaching complete understanding of the part in AI, the vision system can guide the robot to pick through the side hole of the case.