πŸ“‹ Case Study

Vision-Guided Arc Welding Cell for Automotive Chassis Assembly

Thermal distortion-induced seam deviation (>1.8 mm) between stamped aluminum panels; inconsistent joint gap due to fixture wear

πŸ—οΈ Project Overview

Tier-1 supplier’s new EV battery enclosure production line

🎯 Challenge

Thermal distortion-induced seam deviation (>1.8 mm) between stamped aluminum panels; inconsistent joint gap due to fixture wear

πŸ”§ Design Approach

Through-the-arc CCD camera with adaptive thresholding; real-time seam tracking using Hough-transform + Kalman filter fusion; closed-loop joint offset correction via URScript TCP redefinition every 25 mm

πŸ“ Design Diagram

Vision-Guided Arc Welding CellAutomotive Chassis Assembly (Aluminum)Thermal distortion >1.8 mmFixture wear β†’ gap varianceCCD CameraThrough-the-arcAdaptive thresholdingHough + KalmanLatency: 9.3 ms<12 ms budgetURScript TCPΞ”z = 0.62 Γ— gap_widthRedefine every 25 mmSeam imageFeature extractionWeld headJoint offset correctionDistorted seam25 mmΞ”z

AI-generated project design illustration

πŸ“ Key Calculations

Seam Tracking Latency Budget

Exposure + Processing + Comms < 12 ms
Result: 9.3 ms
Prevents arc blowout during high-speed welds

Joint Gap Compensation Factor

Ξ”z = k Γ— gap_width
Result: k = 0.62
Maintains consistent penetration depth across 0–3.2 mm gaps

πŸ“Š Results

Weld rejection rate reduced from 8.3% to 0.4%; cycle time shortened by 11.5 sec/part; post-weld inspection pass rate increased to 99.97%

πŸ’‘ Lessons Learned

  • β€’Through-the-arc imaging requires spectral filtering at 650Β±10 nm
  • β€’Kalman prediction improved tracking continuity during brief occlusions
  • β€’URScript-based TCP update avoids costly PLC-level integration

βœ… Key Takeaways

  • 1Through-the-arc imaging requires spectral filtering at 650Β±10 nm
  • 2Kalman prediction improved tracking continuity during brief occlusions
  • 3URScript-based TCP update avoids costly PLC-level integration