The Challenge

Sorting colored gemstones is traditionally a slow, manual process due to the irregular shapes and varying optical properties of the stones. The goal was to automate this classification by weight and optical quality without damaging the high-value material.

The Engineering Solution

We developed a hybrid automation architecture that bridges standard industrial robotics with custom-engineered material handling units.

  • Robotic Integration: We deployed an SHR3-600 SCARA robot to handle the high-speed pick-and-place operations, utilizing a custom-designed pneumatic gripper system to handle delicate stones.
  • Proprietary Feeding: To solve the issue of irregular stone flow, we engineered and manufactured an in-house rotary drum feeding system that works in tandem with a standard conveyor belt to separate stones for the camera.
  • Computer Vision (CV): The vision unit utilizes proprietary algorithms to analyze 3D shape, volume, color, and internal inclusions (transparency), sending coordinates directly to the robot.

Technical Scope

  • Control Architecture: A multi-threaded Finite-State Machine (FSM) synchronizes the scales, camera tracking, and robot actuation via Modbus/TCP.
  • Precision Weighing: Integration of an array of six precision mini-scales to categorize stones by carat weight alongside optical data.
  • Electrical Design: Custom PCB and relay integration for solenoid valves, vacuum systems, and stepper motor drivers to control the feeding mechanism.