Increasing the Robustness of Random Bin Picking by Avoiding Grasps of Entangled Workpieces

Moosmann, Marius; Spenrath, Felix; Kleeberger, Kilian; Khalid, Muhammad Usman; Mönnig, Manuel; Rosport, Johannes; Bormann, Richard

doi:10.1016/j.procir.2020.03.082

Cited by 25 publications

(12 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To validate the effort of separation in simulation, we propose a method considering the displacement of the entangled workpieces. First of all, we use the cropped depth image from [1] which is oriented along a certain workpiece coordinate axis and span a Fig. 1.…”

Section: Methods To Validate the Effort Of Separation In Simulationmentioning

confidence: 99%

“…Entangled workpiece situations within bin picking can cause increasing cycle times and a loss of robustness in production. To cope with these situations, we previously developed a method to detect and avoid entangled workpiece situations in a pile [1]. If however, there are no unentangled workpieces left, another method is needed to separate those entanglements to be able to empty the bin completely.…”

Section: Introductionmentioning

confidence: 99%

“…Spenrath et al [4] proposed a method with fixed approach and depart poses that are checked for collisions with the point cloud. Considering coping with entangled workpiece situations in a pile, Moosmann et al [1] focused on a model-based approach to detect those situations and to avoid them while bin picking. Matsumura et al [5] proposed detecting gripping points in a pile that can lead to a failed grasp.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Using Deep Neural Networks to Separate Entangled Workpieces in Random Bin Picking

Moosmann

Spenrath

Mönnig

et al. 2021

Advances in Automotive Production Technology – Theory and Application

Self Cite

View full text Add to dashboard Cite

Entanglements can cause robots to pick multiple parts within random bin picking applications. Previous approaches cope with this problem by shaking the gripped workpiece above the bin. However, these methods increase the cycle time and may decrease the robustness of the application. Therefore we propose a new method to separate entangled workpiece situations by using deep supervised learning. To generate annotated training data for a convolutional neural network we set up a simulation scene. In this scene, bins are filled with different amounts of sorted workpieces in several entangled situations. Each workpiece is then moved into different directions to path poses which are evenly distributed along the surface of a hemisphere. The emerging dataset consists of cropped depth images of entangled workpiece situations and several path poses. A serial connection of convolutional neural networks is trained on this dataset and proposes a sequence of poses yielding the general departure path. Finally, the performance of this method is validated on simulated data. To the best of our knowledge, our proposed method is the first systematic approach to find the best extraction strategy to separate entangled workpieces in a pile while decreasing the effective cycle time for gripping entangled workpieces and increasing the robustness significantly.

show abstract

Section: Methods To Validate the Effort Of Separation In Simulationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Using Deep Neural Networks to Separate Entangled Workpieces in Random Bin Picking

Moosmann

Spenrath

Mönnig

et al. 2021

Advances in Automotive Production Technology – Theory and Application

Self Cite

View full text Add to dashboard Cite

show abstract

“…Furthermore, our method cannot do pre-grasp manipulations on the objects to change their poses in order to more robustly grasp them. While PQ-Net can avoid entangled object situations, we do not propose a solution to unhook very complex object geometries, for which no general solution has been proposed so far [44], [45].…”

Section: B Limitationsmentioning

confidence: 99%

Transferring Experience from Simulation to the Real World for Precise Pick-And-Place Tasks in Highly Cluttered Scenes

Kleeberger¹,

Völk²,

Moosmann³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

In this paper, we introduce a novel learning-based approach for grasping known rigid objects in highly cluttered scenes and precisely placing them based on depth images. Our Placement Quality Network (PQ-Net) estimates the object pose and the quality for each automatically generated grasp pose for multiple objects simultaneously at 92 fps in a single forward pass of a neural network. All grasping and placement trials are executed in a physics simulation and the gained experience is transferred to the real world using domain randomization. We demonstrate that our policy successfully transfers to the real world. PQ-Net outperforms other model-free approaches in terms of grasping success rate and automatically scales to new objects of arbitrary symmetry without any human intervention.

show abstract

“…Az alkatrészek megfogáshoz gyakran univerzális két-három-négyujjas megfogókat használnak [6], de alkalmaznak speciális megfogószerkezeteket is, melyek például pneumatikus vagy mágneses elven működnek [7]. Továbbá léteznek olyan rendszerek is, amelyek több megfogó használatát teszik lehetővé, ezzel biztosítva több különböző alkatrész hatékony mozgatását [8,9].…”

unclassified

Rendezetlen alkatrészek rakodási folyamatának fejlesztése

Magyar¹,

Sebestyén²,

Magyar³

et al. 2021

eis

View full text Add to dashboard Cite

Ebben a cikkben egy automatizált rakodó (bin-picking) rendszer kialakításának, illetve tovább fejlesztésének lépéseit fogjuk bemutatni. A rendszernek a feladata, hogy a rendezetlenül szállított tengelycsonkokat egy robot rendezetten a palettára helyezze. Az automatizált rendszer két elemét fogjuk részletesen bemutatni: a robotprogramot és forgató asztal működését. Kitérünk a robotprogram fő- és alprogramjai közötti összefüggésekre és a működéshez megfelelő struktúra megvalósítására. A forgató asztal működésének fejlesztését több lépésen keresztül mutatjuk be, szisztematikus mérésekkel igazolva az eredményeket. A rendszer működőképességét azóta több ezer kipakolás is igazolja.

show abstract

Increasing the Robustness of Random Bin Picking by Avoiding Grasps of Entangled Workpieces

Cited by 25 publications

References 12 publications

Using Deep Neural Networks to Separate Entangled Workpieces in Random Bin Picking

Using Deep Neural Networks to Separate Entangled Workpieces in Random Bin Picking

Transferring Experience from Simulation to the Real World for Precise Pick-And-Place Tasks in Highly Cluttered Scenes

Rendezetlen alkatrészek rakodási folyamatának fejlesztése

Contact Info

Product

Resources

About