Case | How 3D Vision Solves the Flexible Assembly Challenge of Air Conditioner Compressors?

On highly automated home appliance production lines, the assembly of outdoor AC units has always been a critical link. Among them, the precise and efficient assembly of compressors directly affects the overall product quality and production efficiency. However, this step faces multiple challenges such as reflective surfaces, variable incoming part positions, and high cycle time requirements.

Today, we will delve into a real-world retrofit project to see how advanced 3D vision technology empowers traditional production lines, enabling intelligent upgrades for compressor assembly.

Project Challenges

The retrofit production line initially faced three major difficulties in the compressor feeding process: - Difficult positioning: The compressor surface is mostly deep black with reflections, making it hard for traditional 2D vision or manual positioning to stably and accurately identify its precise location and orientation. - Tight cycle time: The designed cycle time is as high as 9 seconds per unit; any delay would cause line congestion, requiring the vision system to complete recognition and guidance in an extremely short time. - Low flexibility: The production line needs to accommodate multiple specifications of air conditioner compressors with varying sizes and shapes, demanding strong adaptability from the system.

Solution

To address these pain points, a fixed-mounted 3D laser galvanometer camera was introduced. This camera acts like a "high-hanging fiery eyes" installed for the robot, overlooking the entire scene and providing precise guidance.

The system workflow is as follows: - Vision positioning: The 3D camera is fixed above two assembly stations. When the roller conveyor delivers the compressor pallet, the camera quickly scans the pallet to obtain high-precision 3D point cloud data. - Intelligent analysis: Built-in self-developed algorithms process the point cloud, unaffected by the black reflective surface, accurately calculating the 3D coordinates and rotation angle of each compressor. - Robot grasping: The coordinate data is sent to the robot in real-time, which precisely grasps the compressor and transfers it to the assembly position. - Precise assembly: The robot accurately places the compressor onto the positioned chassis fixing bolts, completing the assembly.

Technical Highlights

In this project, the 3D vision system demonstrated outstanding technical performance: - Large field of view combined with high precision: The camera achieves a large field of view and ±1.2mm recognition accuracy within a working distance of 1.2m to 3.0m, perfectly meeting the multi-target positioning needs of large workpieces. - Ultimate production cycle time: From camera capture to algorithm output coordinates, the entire process is compressed to within a few seconds, helping the line achieve a stable cycle time of ≤9 seconds. - 99.9% operational stability: The system features strong environmental adaptability, unaffected by factory lighting variations, and effectively overcomes the black surface and reflection interference of compressors, ensuring continuous and stable production.

This project has successfully validated the maturity of 3D vision technology in high-cycle, demanding industrial applications. Through stable and reliable recognition and guidance, it not only solves the precise positioning problem in compressor assembly but also provides a replicable successful paradigm for the intelligent upgrade of home appliance and other manufacturing industries. In the future, with continuous technological iteration and deepening application scenarios, 3D vision will undoubtedly inject more vitality into "Intelligent Manufacturing in China."