Intelligent Collaborative Robots Augmented with Vision Perception for Flexible Manufacturing System

Document Type

Article

Publication Date

7-2025

Abstract

To accurately detect and localize objects within a robot’s environment in real time, while also enabling to generate instantaneous and effective path planning for part handling, is a critical advancement in the realm of collaborative robots, particularly within the context of flexible manufacturing systems (FMS). Though existing literature provides several methods for 3D object detection, few have considered the unique demands and dynamics of a collaborative robotic environment within FMS. This paper attempts to integrate a vision perception system into a collaborative robotic system in order to address the aforementioned challenge. In particular, using the advanced capabilities of Franka Emika Panda robots and the detailed environmental data captured by a 3D camera, a system for real-time object pose detection and computation is developed. This achievement enables swift and accurate decision-making, reduces the need for reliance on preprogrammed tasks, and ultimately enhances the adaptability of the robots within the FMS. The proposed software algorithm not only detects and localizes the position and orientation of a part but also executes path planning to accurately pick up and place the detected part. Preliminary testing, which includes tasks involving part pickup and placement, suggests that the integration of vision perception systems can substantially augment the functionality, efficiency, and adaptability of the collaborative robots within an FMS environment. This paper contributes to the evolving conversation around manufacturing automation and lays the foundation for future research and innovation in this arena.

Link to Full Text

Share

COinS