A close-up, high-angle view of a large pile of randomly assorted, heavy laminated steel parts, demonstrating a complex challenge for robotic bin picking and the need for a precise AI vision solution.

Bin Picking Steel Parts Using AI

Case Overview

Motion Technique logo - India-based industrial automation systems integrator.

Customer: Motion Technique (systems integrator for Siemens)

Location: India

Industry: Electronics / Smart Manufacturing

Solution: AccuPick

Objectives:

  • Achieve accurate recognition of laminated steel component position/orientation
  • Enable collision-free bin picking in deep containers
  • Meet strict cycle time requirements for switchgear manufacturing

Results:

  • 100% collision-free picking with long gripper in deep bin
  • Achieved 6-second cycle time for pick-and-place
  • Improved accuracy in component orientation recognition

Testimonial:

“We are glad that we chose AccuPick for one of the most challenging bin picking projects that came our way. We believe no other bin picking solution would have been able to execute this project of such high complexity.”

Uday Mali, Director, Motion Technique

Uday Mali, Director, Motion Technique

The Customer

Motion Technique, a systems integrator based in India, partnered with Solomon to enhance automation at Siemens’ switchgear manufacturing facility. Siemens, a global leader in electrical engineering and manufacturing, required a highly precise and efficient bin picking solution to handle complex laminated steel components in their production process.

The Challenge

Overcoming Precision and Collision Risks in Production

Siemens needed a reliable way to pick small, heavy laminated steel components from deep bins without collisions, while meeting strict cycle time requirements. The parts’ complex shapes, random orientations, and tendency to tangle made CAD-based localization ineffective, creating risks to productivity, quality, and operational efficiency.

The Solution

Customised Automation for Complex Part Handling

To address component picking challenges from deep bins, Motion Technique deployed Solomon’s AccuPick AI-based bin picking system together with a SolScan 2.3 MP scanner, integrated with a Universal Robots UR10 cobot. This setup enabled precise orientation detection, collision-free operation, and seamless component placement via a custom slider, fully meeting Siemens’ accuracy and speed requirements.

Screenshot of the AccuPick AI vision interface, showing the deep learning annotation and training process for robotic bin picking of switchgear components.
AccuPick software interface image annotation to train deep learning AI model for robotic bin picking
Robotic bin picking of randomly oriented steel switchgear parts from a metal bin in a manufacturing cell, showcasing industrial AI vision automation.
AccuPick integrates with UR cobots to automate handling of randomly oriented steel switchgear parts

The Results

Achieved 6-second cycle time from scanning to placement
100% collision-free operation with long gripper in deep bin
Fully met Siemens’ operational requirements for speed and accuracy

Application Video