Bin Picking Reflective Parts

Context & Aim

The goal is to find a suitable technology capable of identification and location of randomly orientated complex shaped parts with different surface finishes to enable a robot system to accurately and repeatably locate and orientate the part for robotic pick and place applications.


Vision guided robotic (VGR) solutions using 3D imaging systems allow robots to locate and pick up parts from varying locations reducing the need for expensive orientation, enabling greater production flexibility, allowing for more tasks to be automated. The challenges presented by the application:


  • Highly Reflective Surfaces
  • Complex Shapes
  • Component obscured by tray walls
  • Different surface finishes
  • Tray Variability


IMR are taking 2 separate approaches and this includes:


Background research on vision systems to down select technology approaches that could be investigated in IMR.


Assessment of 3rd Party technologies:

  • Internal Technology assessment
  • 2D Vision with CAD matching
  • Deflectrometery from Viztronics
  • Possible external Vendors
  • IDS
  • Euclid Labs


Highly reflective parts provide considerable problems for vision system scanning and is a challenge that has been set to IMR by the REDF consortium. Adding in Bin Picking of randomly located parts and complex shapes increased the challenge to make it a worthy proposal for IMR research.

Technical specifications

Part location

Must be able to define the location of the part in 6 axis X,Y,Z and Roll, Pitch Yaw

Surface finish

Be able to locate the parts provided by DePuy with Both highly reflective and Matt finishes

X,Y,Z Location

Be able to accurately and repeatably locate the part in X,Y,Z to +/- 3mm


Ken Horan
Senior Manager Robotics & Automation


Anum Rehman
Robotics Researcher

Francis O’Farrell
Senior Technician


Robotics and Automation