3D Machine Vision Techniques

FlexiPick Random Bin Picking

Random Robotic Bin Picking to Improve Productivity & Throughput

The Application

A global Tier-1 automotive parts manufacturer requires a robot pick and place solution to automate the placement of randomly placed car axles from a storage bin into a flange that leads into a conveyor belt using 3D machine vision techniques.
Robot pick and place
Wheel axle shafts connected to the differential

The Challenge

The customer originally had a 2D vision system and robot pick and place for the bin picking application from a different integrator. Due to poor reliability, accuracy and inadequate performance, this robot pick and place system was never implemented into a production environment. Moreover, manual loading of the axle is also impractical due to the heavy weight and large dimension, which poses a serious health and safety risk for the operator.

 

To help overcome this challenge, Bluewrist proposed our FlexiPick automated random bin picking solution that uses a 3D scanner for greatly improved parts recognition, collision prevention and fast processing.

The Solution – 3D Machine Vision Techniques

The manufacturer required a 100% inline picking solution to eliminate the manual handling and loading of the car axles onto to the conveyor for further processing.

 

In order to meet the customer’s requirement and address the challenge they face, Bluewrist proposed our advanced bin-picking solution, Bluewrist FlexiPick system, which uses a 3D snapshot sensor positioned top of the axle storage bin. The 3D sensor generates point cloud of the parts in the bin which is then streamed to comXtream, a Bluewrist industrial communication software for additional processing.

 

Bluewrist FlexiPick system receives the point cloud data from comXtream, which would identify the best part for picking. The robot will be commanded to pick up the part using a gripper specifically designed for this project.

Case Study – FlexiPick Random Bin Picking
Wheel axle shafts placed randomly in the bin
3D machine vision techniques
Reliable detection of misplaced parts in bin and collision check

The Results

  • Our solution is able to identify parts in the bin every 3 to 8 seconds depending factors such as bin fullness or part orientation with 95% plus success rate.

 

  • Our 3D vision based bin picking solution is highly robust and reliable in a production environment and is not affected by changes in ambient lighting conditions or temperature

 

  • Bluewrist software solution is hardware agnostic. Customer is able to reuse existing robot, PLCs and conveyor hardware that greatly reduced deployment time.

Next Steps

After the successful deployment of the first bin-picking solution, the customer will adopt similar systems in several other production lines. This will lead to significant increase in productivity and bottleneck reduction through the elimination of manual material handling.

Bluewrist Advantage

A high performance and reliable random bin picking solution requires seamless integration of software, 3D vision sensors and robotic guidance. Powered by custom software that is precisely calibrated to the physical characteristics of the part and tailored to meet a manufacturers’cycle time requirements, FlexiPick seamlessly integrates into the customer’s existing production process.

Please contact us for more information.

flexipick_axle_picking
FlexiPick random picking of wheel axles