FANUC industrial robot picking a metal ring from a deep metal bin using 3D machine vision
Machining · Bin Picking

Automated Bin Picking of Metal Rings

3D machine vision and a FANUC robot turn chaotically stacked rings into a precise, uninterrupted feed for a cold-rolling machine.

Delivered Project

Industries

  • Machining
  • Metal components

Tech Stack

  • FANUC 6-axis robot
  • 3D machine vision
  • Chaotic-part algorithm
  • Extended gripper
  • Vertical fixture

Context and challenge

Parts arrive loose in metal bins. The task: extract them one by one and position each precisely for the next machine — with no human intervention.

Chaotic geometry

Parts overlap at random orientations — 2D vision cannot deliver reliable recognition.

Collision risk

The robot works deep inside the bin, millimetres from the walls and neighbouring parts, on every cycle.

Monotony and fatigue

Manually handling thousands of identical rings leads to errors and inconsistent cycle times.

Downstream interruptions

The cold-rolling machine needs an even feed — every pause is lost throughput.

Solution

An integrated cell combining a FANUC robot, a 3D scanning system and a purpose-built gripper — engineered around the real ring geometry and the depth of the bin.

1

3D vision system

Scans the bin on every cycle and builds a fresh 3D point cloud — the foundation of every pick decision.

2

Chaotic-part algorithm

Determines which part is on top and at what angle to grip it, while ruling out collisions before motion begins.

3

FANUC 6-axis robot

Industrial manipulator with the reach for deep bins, programmed for long pick-and-place cycles.

4

Extended gripper

Magnetic or mechanical, engineered to reach the bottom of deep containers without risking the tool.

5

Vertical fixture

Receives each part and aligns it with repeatable precision for the next machine.

6

Central PLC control

Synchronises the vision system, the robot and the downstream operation into one uninterrupted cycle.

How the cycle runs

One cycle — from chaotic bulk inside the bin to a precisely positioned part, ready for the next machine.

Scan

The vision system images the bin and builds a 3D point cloud of every visible part.

Target selection

The software computes the optimal trajectory and picks the part that can be grasped without collision.

Pick

The robot descends into the bin and lifts a single metal ring with its extended specialised gripper.

Place

The part is transferred onto the vertical fixture and aligned precisely for the next operation.

Repeat

The cycle continues autonomously until the bin is empty — with no human intervention.

2–3×

Typical throughput vs manual picking

±0.3 mm

Typical placement precision at handoff

24/7

Unattended operation capable

−70%

Typical labour reduction on the task

Chaotic-parts problems dissolve with intelligent scanning — a slow manual task becomes a fast, reliable automated one.

Have a similar application in mind?

Send us your requirements and we'll outline a concept proposal.