Volt Electric Motors – Vision-Based Quality Control System

Electric motor production involves processes requiring high precision and repeatability. In critical steps such as stator manufacturing, even the smallest error can directly affect motor performance. For years, Volt Electric Motors relied on operator observation for quality control in this critical process.

However:

• Micro defects that cannot be detected by the human eye,
• Fatigue-induced inattentiveness,
• And time-consuming inspection processes adversely affecting quality.

To address these challenges, we set out to develop a fully vision-based automated quality control system.

Our team developed a vision-processing infrastructure integrated into the production line and supported by high-resolution cameras, along with associated analysis software.

Project scope included:

• Industrial cameras suitable for the line were placed in tight spaces
• Custom image processing algorithms built with Python and .NET were developed
• Shape, gap, alignment and structural checks of produced stators were performed instantly
• Defective products were flagged on the line and reported to the operator by the system
• Quality data was archived on a central server and made reportable

Volt Electric Motors faced the challenge that in a complex production flow, deviations could only be detected through manual retrospective checks, causing time loss, quality issues and repeated errors.

The main problems we aimed to solve in this project were:

• Operator-caused defects overlooked by visual inspection
• Inspection speed failing to keep up with production speed
• Off-spec products entering the shipment chain

With the system we developed: • The quality control process was fully automated • Human-factor errors were eliminated • Real-time inspection keeping pace with production was achieved • Quality issues began to be detected during production, not afterward

When the project was completed, measurable and impressive gains were achieved:

• Quality-related defects in stator production dropped by 90%
• Inspection time per unit was reduced to mere seconds
• Customer returns decreased significantly
• The number of defective products leaving production was drastically reduced
• Quality control data became storable digitally for retrospective tracking