Robotic Machine Tending with 3D vision 

Automate machine tending with millimeter accuracy. Reliable 3D vision for reflective metals, fewer errors, faster setup, and stable production performance.

Automating the Toughest
Robotic Machine Tending Challenges

From manual loading to fully automated laser marking - developed by Gasbarre and Pickit 3D, powered by a Zivid 3D camera.

Robotic machine tending automates part loading and unloading using industrial or collaborative robots to increase machine utilization and reduce manual intervention. The robot must position raw or finished parts within tight tolerances while interacting reliably with CNC machines, presses, or other equipment.

Insertion and placement tasks typically require millimeter-level accuracy. Minor deviations can cause misloads, collisions, scrap, or unplanned downtime. Robust 3D vision is therefore a critical subsystem in modern machine tending cells. The camera must deliver accurate, low-noise, and repeatable 3D data across varying materials and lighting conditions to ensure deterministic pick-and-place performance.

Precise Placement & Orientation

3D depth icon
Robots must detect fine orientation details and separate parts reliably in varying lighting conditions.

Reflective and Dark Materials

icon shiny metal parts for surface finishing v3
Shiny or dark parts create distorted or incomplete 3D data for conventional vision systems.

Millimeter-level precision

Robot mounted vision icon-1
Tight tolerances require precise alignment without constant jig adjustments or rework.

Achieve machine tending quality
with Zivid 3D sensing

rusty_bolts_combined v2
standing_metal_cylinders combined v2
Thin metal cylinders
metal_spindle_combined_final
shiny_metal_disks_organized_combined v2
Shiny metal disks - unorganized
rusty_bolts_combined v2
standing_metal_cylinders combined v2
Thin metal cylinders
metal_spindle_combined_final
shiny_metal_disks_organized_combined v2
Shiny metal disks - unorganized
Point cloud examples of machine tending captured with Zivid 3D cameras. Click the image to explore in 3D.

Enable Reliable Pick and Place

  • Clearly separates parts and captures detailed orientation features

  • Handles ordered and unordered parts without pre-sorting

  • High-quality 2D color and stable 3D data in changing light

  • Low noise and minimal distortion for first-time-right grasping

One-motion hook placement powered by Innodura and a Zivid 3D camera.
rusty_bolts_rainbow 50_50

Expand What Your Cell Can Handle

  • Covers metals, plastics, and composite materials

  • Complete point clouds — even on highly reflective surfaces

  • Wide dynamic range (HDR) for extreme contrast scenes in a single capture

  • Fast capture times (~500 ms) on challenging parts

Reduce Setup Time and Manual Adjustments

  • Stationary or robot-mounted configurations for optimal viewing geometry

  • Accurate grasping in a single shot, eliminating the need for realignment jigs.

  • Strong ambient light robustness for reliable factory-floor deployments

  • Zivid presets and infield correction tools to minimize parameter tuning and commissioning time
Kolektor Digital's KoCo robot, equipped with a Zivid 3D camera.

Detecting stacked, reflective sheet metal parts under varying light conditions was a major challenge — Zivid 2+ delivered the best price-performance ratio, with stable, high-resolution 3D data that works even with mirroring.

Matthias Frey, Head of Robotics at J. Schmalz GmbH | Powered by a Zivid camera

Meet our customer team for a technical demo & discussion

Why traditional approaches fail
in real production

Typical Limitations

 

❌ Inconsistent results across varying part geometries and sizes.

❌ Data loss or artifacts on reflective or dark surfaces.

❌ Unstable boundary detection affecting grasp and insertion accuracy.

❌ Fixed mounting reduces flexibility in viewing angle optimization.

❌ Sensitivity to ambient light and long-term drift.

 

 

Zivid solutions

 

✅ High-accuracy 3D data enabling millimeter-level placement. Learn more →

✅ Reliable capture of specular and low-reflectivity materials. Learn more →

✅ Stable pose estimation without costly realignment jigs. Learn more →

✅ Flexible mounting strategies for optimal imaging geometry. Learn more → 

 ✅ Proven stability in industrial production environments. Learn more →  

Machine tending automation
in the real world

Circu Li-ion | Automated battery upcycling solution
circulion
Robot-mounted

Circu Li-ion | Automated battery upcycling solution

Circu Li-ion's automated solution enables companies to significantly reduce CO2 emissions and increase the true value of each battery.
Innodura | Complex parts hanging solution for assembly lines
Untitled (1144 × 614px)-2-1
Robot-mounted

Innodura | Complex parts hanging solution for assembly lines

Hanging parts for processing is a task that has been primarily carried out manually, and automating it has proven to be a real challenge.
Kolektor | Envisioning robot workers
Kolektor-KoCo-production-setup-Zivid-Two-Machine-Tending-1
Robot-mounted

Kolektor | Envisioning robot workers

Kolektor Digital needed a 3D machine vision camera with human-like capability.

Start automating machine tending
with Zivid 3D vision

Explore machine tending point clouds

Explore machine tending point clouds

Book a free demo
with us

Book a free demo
with us

Get pricing for
your application

Get pricing for
your application

Zivid’s 3D machine vision technology powers smarter factories and warehouse automation. Our market-leading 3D color cameras and software enable customers to increase efficiency and productivity across industrial applications such as robotic bin-picking, piece-picking, assembly, inspection, and depalletization.