TY - JOUR
T1 - Learning to place objects onto flat surfaces in upright orientations
AU - Newbury, Rhys
AU - He, Kerry
AU - Cosgun, Akansel
AU - Drummond, Tom
PY - 2021/7
Y1 - 2021/7
N2 - We study the problem of placing a grasped object on an empty flat surface in an upright orientation, such as placing a cup on its bottom rather than on its side. We aim to find the required object rotation such that when the gripper is opened after the object makes contact with the surface, the object would be stably placed in the upright orientation. We iteratively use two neural networks. At every iteration, we use a convolutional neural network to estimate the required object rotation, which is executed by the robot, and then a separate convolutional neural network to estimate the quality of a placement in its current orientation. Our approach places previously unseen objects in upright orientations with a success rate of 98.1% in free space and 90.3% with a simulated robotic arm, using a dataset of 50 everyday objects in simulation experiments. Real-world experiments were performed, which achieved an 88.0% success rate, which serves as a proof-of-concept for direct sim-to-real transfer.
AB - We study the problem of placing a grasped object on an empty flat surface in an upright orientation, such as placing a cup on its bottom rather than on its side. We aim to find the required object rotation such that when the gripper is opened after the object makes contact with the surface, the object would be stably placed in the upright orientation. We iteratively use two neural networks. At every iteration, we use a convolutional neural network to estimate the required object rotation, which is executed by the robot, and then a separate convolutional neural network to estimate the quality of a placement in its current orientation. Our approach places previously unseen objects in upright orientations with a success rate of 98.1% in free space and 90.3% with a simulated robotic arm, using a dataset of 50 everyday objects in simulation experiments. Real-world experiments were performed, which achieved an 88.0% success rate, which serves as a proof-of-concept for direct sim-to-real transfer.
KW - Deep learning in grasping and manipulation
KW - perception for grasping and manipulation
UR - http://www.scopus.com/inward/record.url?scp=85103301945&partnerID=8YFLogxK
U2 - 10.1109/LRA.2021.3068122
DO - 10.1109/LRA.2021.3068122
M3 - Article
AN - SCOPUS:85103301945
SN - 2377-3766
VL - 6
SP - 4377
EP - 4384
JO - IEEE Robotics and Automation Letters
JF - IEEE Robotics and Automation Letters
IS - 3
ER -