Andre Yuji Yasutomi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Andre Yuji Yasutomi has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Hideo Enoki
Robotics Research Department, Hitachi Ltd., Hitachinaka, Ibaraki Japan 312-0034
1 shared publication - Hideyuki Ichiwara
Research and Development Group, Hitachi Ltd., Ibaraki, Japan
1 shared publication - Hiroki Mori
Department of Intermedia Art and Science, School of Fundamental Science and Engineering, Waseda University, Tokyo, Japan
1 shared publication - Hiroshi Ito
Department of Intermedia Art and Science Waseda University Hyogo Tokyo Japan
1 shared publication - Tetsuya Ogata
Department of Intermedia Art and Science, Waseda University, Tokyo, Japan
1 shared publication
Latest Publications
- 8 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 8 (3) , 1834-1841
Visual Spatial Attention and Proprioceptive Data-Driven Reinforcement Learning for Robust Peg-in-Hole Task Under Variable Conditions
Anchor-bolt insertion is a peg-in-hole task performed in the construction field for holes in concrete. Efforts have been made to automate this task, but the variable lighting and hole surface conditions, as well as the requirements for short setup and task execution time make the automation challenging. In this study, we introduce a vision and proprioceptive data-driven robot control model for this task that is robust to challenging lighting and hole surface conditions. This model consists of a spatial attention point network (SAP) and a deep reinforcement learning (DRL) policy that are trained jointly end-to-end to control the robot. The model is trained in an offline manner, with a sample-efficient framework designed to reduce training time and minimize the reality gap when transferring the model to the physical world. Through evaluations with an industrial robot performing the task in 12 unknown holes, starting from 16 different initial positions, and under three different lighting conditions (two with misleading shadows), we demonstrate that SAP can generate relevant attention points of the image even in challenging lighting conditions. We also show that the proposed model enables task execution with higher success rate and shorter task completion time than various baselines. Due to the proposed model's high effectiveness even in severe lighting, initial positions, and hole conditions, and the offline training framework's high sample-efficiency and short training time, this approach can be easily applied to construction.
- 8 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 8 (3) , 1834-1841
Visual Spatial Attention and Proprioceptive Data-Driven Reinforcement Learning for Robust Peg-in-Hole Task Under Variable Conditions
Anchor-bolt insertion is a peg-in-hole task performed in the construction field for holes in concrete. Efforts have been made to automate this task, but the variable lighting and hole surface conditions, as well as the requirements for short setup and task execution time make the automation challenging. In this study, we introduce a vision and proprioceptive data-driven robot control model for this task that is robust to challenging lighting and hole surface conditions. This model consists of a spatial attention point network (SAP) and a deep reinforcement learning (DRL) policy that are trained jointly end-to-end to control the robot. The model is trained in an offline manner, with a sample-efficient framework designed to reduce training time and minimize the reality gap when transferring the model to the physical world. Through evaluations with an industrial robot performing the task in 12 unknown holes, starting from 16 different initial positions, and under three different lighting conditions (two with misleading shadows), we demonstrate that SAP can generate relevant attention points of the image even in challenging lighting conditions. We also show that the proposed model enables task execution with higher success rate and shorter task completion time than various baselines. Due to the proposed model's high effectiveness even in severe lighting, initial positions, and hole conditions, and the offline training framework's high sample-efficiency and short training time, this approach can be easily applied to construction.
- 18 February 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 5 (2) , 2921-2928
Localization of Inspection Device Along Belt Conveyors With Multiple Branches Using Deep Neural Networks
Regular inspections of belt conveyors are required to prevent the damage of transported objects. Nevertheless, inspections can be troublesome for belt conveyors composed of a plurality of belt lines with multiple branches. To improve the inspection process, an inspection device composed of an inertial measurement unit (IMU) inside a transported object was introduced in our previous study jointly with an algorithm to detect anomalies in the joints of the belt lines. When belt conveying this device for inspection, however, it is required to not only detect the anomaly but also know its position. This study presents a novel method to estimate the position of the inspection device along the belt conveyor using a deep neural network (DNN). The DNN uses the IMU data to detect seven types of features (passage through five types of joints, device stoppage and regular transport), which, when matched to data on a belt conveyor position database, can correctly be translated into positions. Additionally, the proposed method enables the detection of changes of routes along the belt conveyor that occur when a belt line branches into two output streams. To enhance the DNN feature detections, two original algorithms for DNN output post-processing are also introduced. Experiments with a complex belt conveyor demonstrate this method can successfully detect the position of the inspected device more accurately and more cost-effectively than conventional methods.
- 18 February 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 5 (2) , 2921-2928
Localization of Inspection Device Along Belt Conveyors With Multiple Branches Using Deep Neural Networks
Regular inspections of belt conveyors are required to prevent the damage of transported objects. Nevertheless, inspections can be troublesome for belt conveyors composed of a plurality of belt lines with multiple branches. To improve the inspection process, an inspection device composed of an inertial measurement unit (IMU) inside a transported object was introduced in our previous study jointly with an algorithm to detect anomalies in the joints of the belt lines. When belt conveying this device for inspection, however, it is required to not only detect the anomaly but also know its position. This study presents a novel method to estimate the position of the inspection device along the belt conveyor using a deep neural network (DNN). The DNN uses the IMU data to detect seven types of features (passage through five types of joints, device stoppage and regular transport), which, when matched to data on a belt conveyor position database, can correctly be translated into positions. Additionally, the proposed method enables the detection of changes of routes along the belt conveyor that occur when a belt line branches into two output streams. To enhance the DNN feature detections, two original algorithms for DNN output post-processing are also introduced. Experiments with a complex belt conveyor demonstrate this method can successfully detect the position of the inspected device more accurately and more cost-effectively than conventional methods.