Automating CNC Machine Tending with 3D Robotic Bin Picking
Case Overview
Industry: Manufacturing / Metal Processing
Solution: AccuPick
A CNC polishing process used robotic bin picking to identify and load unpolished metal workpieces into CNC machines. The application involved picking distances exceeding 1,250 mm and required positioning accuracy within ±1 mm to ensure reliable gripping and prevent collisions during machine loading.
The Case
CNC Machine Loading with Robotic Bin Picking
A multinational chemical glass manufacturer needed to automate the loading of unpolished metal workpieces into CNC machines.
The parts were stored in a bin with varying positions and orientations. Before polishing, the automation system needed to locate an accessible workpiece, calculate its position and orientation, and provide the robot with the coordinates required for picking and placement.
Accurate CNC machine tending was necessary because incorrect gripping or placement could interrupt the loading cycle or cause a collision between the robot, workpiece, and machine.
The Challenge
Reliable Detection of Unpolished Metal Workpieces
The workpieces had unpolished metal surfaces that were difficult for traditional vision systems to identify consistently. Surface characteristics and random part orientations affected the system’s ability to generate stable picking positions.
The vision system also needed to detect parts at picking distances exceeding 1,250 mm within the bin. This large working range increased the difficulty of maintaining the required positioning accuracy.
The application specified picking accuracy within ±1 mm. Deviations outside this tolerance could result in incomplete gripping, incorrect placement, or collisions during CNC machine loading.
The Solution
3D Vision-Guided CNC Machine Tending
The customer deployed AccuPick to automate CNC machine tending. This AI-based robotic bin picking system enabled robots to identify, locate, and pick randomly positioned workpieces from bins.
The SolScan industrial 3D camera captured the workpiece arrangement inside the bin and generated a 3D point cloud in 1.5 seconds. AccuPick processed the 3D data to locate target workpieces and calculate six-dimensional object coordinates containing position and orientation information.
The system generated robot picking points and transferred the required coordinates to the robot for part retrieval and placement into the CNC machine. Standard PLC communication protocols supported integration with the customer’s existing automation system.
Workpiece Detection and Robot Guidance
During each loading cycle, AccuPick analyzed the 3D vision data to identify an accessible workpiece and provide the robot with the coordinates required for picking and loading into the CNC polishing machine.