Research on robot automatic assembly system based on automatic programming

This paper introduces an innovative automated robotic assembly system for intelligent manufacturing, designed to minimize setup time and costs by reducing the need for frequent task reassignments in robotic operations. The proposed system combines two core elements: robots learning assembly tasks through human demonstrations and embodying these learned tasks as robot skills. Initially, robots observe human assembly operations, which are formalized into human assembly scripts. These scripts are then converted into robot-specific assembly tasks through workspace modeling and action planning, supported by simulation redirection, enabling the execution of a range of tasks. The system equips robots with pre-trained skills, enabling them to autonomously perform complex tasks that involve inherent uncertainty and variability. To validate the approach, a prototype assembly system was developed for power circuit breakers and electronic set-top boxes. The experimental results confirm that the proposed system is not only feasible but also significantly reduces the need for human intervention in automated robotic assembly, highlighting its potential to optimize manufacturing processes. This approach offers a promising solution to enhance the efficiency, flexibility, and cost-effectiveness of robotic automation in production environments, making it suitable for a wide range of industrial applications.