Enhancing Industrial Robots with Machine Vision and Low-Code Platforms: A Closed-Loop Approach

Traditional industrial robot control systems often rely on pre-programmed coordinates, limiting adaptability in dynamic environments. To overcome this rigidity, we present a visual-enhanced low-code platform that integrates machine vision for flexible task orchestration. The system uses a depth-sensing camera to capture real-time environmental data, combined with advanced object recognition and grasp algorithms. Experimental validation shows 93% success in grasping randomly placed objects and 95% precision in liquid pouring tasks (average error: 1.2 mm), eliminating dependency on fixed coordinates. This work demonstrates a practical closed-loop control framework for intelligent manufacturing, significantly enhancing adaptability in unstructured scenarios.