University of Twente Student Theses

Login

Design of a No Reference Random Bin Picking System For Metal Plate Parts

Berg, C. van den (2023) Design of a No Reference Random Bin Picking System For Metal Plate Parts.

Full text not available from this repository.

Full Text Status:Access to this publication is restricted
Embargo date:24 August 2028
Abstract:By the production of custom cut plate material parts post processing is often needed. These production lines often use accurately stacked pallets as input. Labor and time could be saved if randomly filled bins could be used as an input of the production line. This thesis describes the design of a bin picking solution for randomly placed metal plate material parts in a bin. The shapes of the parts are not known beforehand. The bin with parts is scanned with a Structured Light Sensor (SLS) to form a point cloud. Research is done on determining how to process the point cloud to detect the parts and to form collision objects. A custom path planning algorithm is created to extract the parts out of the bin with a tall magnet gripper. Besides these main contributions time is spend on handling parts and a method to stack them on an output pallet. The system is tested with within a simulation environment to measure the calcula- tion performance. The system is also tested with an actual robot to measure the pick rate and pick time. The results of the tests with the actual robot setup delivered a pick success rate of 91.42%. The average time to calculation a pick up path is 2.91 seconds per part. The time needed to pick a product and stack it onto the output pallet takes 34.88 seconds on average.
Item Type:Essay (Master)
Clients:
Teqram BV, Zwolle, NL
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Embedded Systems MSc (60331)
Link to this item:https://purl.utwente.nl/essays/97266
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page