Motion switching with sensory and instruction signals by designing dynamical systems using deep neural network

Kanata Suzuki, Hiroki Mori, Tetsuya Ogata

Research output: Contribution to journalArticle

Abstract

To ensure that a robot is able to accomplish an extensive range of tasks, it is necessary to achieve a flexible combination of multiple behaviors. This is because the design of task motions suited to each situation would become increasingly difficult as the number of situations and the types of tasks performed by them increase. To handle the switching and combination of multiple behaviors, we propose a method to design dynamical systems based on point attractors that accept (i) 'instruction signals' for instruction-driven switching. We incorporate the (ii) 'instruction phase' to form a point attractor and divide the target task into multiple subtasks. By forming an instruction phase that consists of point attractors, the model embeds a subtask in the form of trajectory dynamics that can be manipulated using sensory and instruction signals. Our model comprises two deep neural networks: A convolutional autoencoder and a multiple time-scale recurrent neural network. In this study, we apply the proposed method to manipulate soft materials. To evaluate our model, we design a cloth-folding task that consists of four subtasks and three patterns of instruction signals, which indicate the direction of motion. The results depict that the robot can perform the required task by combining subtasks based on sensory and instruction signals. And, our model determined the relations among these signals using its internal dynamics.

Original languageEnglish
Article number8405582
Pages (from-to)3481-3488
Number of pages8
JournalIEEE Robotics and Automation Letters
Volume3
Issue number4
DOIs
Publication statusPublished - 2018 Oct 1

Fingerprint

Dynamical systems
Dynamical system
Neural Networks
Motion
Attractor
Robots
Robot
Recurrent neural networks
Multiple Time Scales
Recurrent Neural Networks
Folding
Model
Trajectories
Divides
Deep neural networks
Trajectory
Internal
Target
Necessary
Evaluate

Keywords

  • AI-based methods
  • Deep learning in robotics and automation
  • humanoid robots

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Human-Computer Interaction
  • Biomedical Engineering
  • Mechanical Engineering
  • Control and Optimization
  • Artificial Intelligence
  • Computer Science Applications
  • Computer Vision and Pattern Recognition

Cite this

Motion switching with sensory and instruction signals by designing dynamical systems using deep neural network. / Suzuki, Kanata; Mori, Hiroki; Ogata, Tetsuya.

In: IEEE Robotics and Automation Letters, Vol. 3, No. 4, 8405582, 01.10.2018, p. 3481-3488.

Research output: Contribution to journalArticle

@article{857fdeb4d6fb4650bae87a34cffd6754,
title = "Motion switching with sensory and instruction signals by designing dynamical systems using deep neural network",
abstract = "To ensure that a robot is able to accomplish an extensive range of tasks, it is necessary to achieve a flexible combination of multiple behaviors. This is because the design of task motions suited to each situation would become increasingly difficult as the number of situations and the types of tasks performed by them increase. To handle the switching and combination of multiple behaviors, we propose a method to design dynamical systems based on point attractors that accept (i) 'instruction signals' for instruction-driven switching. We incorporate the (ii) 'instruction phase' to form a point attractor and divide the target task into multiple subtasks. By forming an instruction phase that consists of point attractors, the model embeds a subtask in the form of trajectory dynamics that can be manipulated using sensory and instruction signals. Our model comprises two deep neural networks: A convolutional autoencoder and a multiple time-scale recurrent neural network. In this study, we apply the proposed method to manipulate soft materials. To evaluate our model, we design a cloth-folding task that consists of four subtasks and three patterns of instruction signals, which indicate the direction of motion. The results depict that the robot can perform the required task by combining subtasks based on sensory and instruction signals. And, our model determined the relations among these signals using its internal dynamics.",
keywords = "AI-based methods, Deep learning in robotics and automation, humanoid robots",
author = "Kanata Suzuki and Hiroki Mori and Tetsuya Ogata",
year = "2018",
month = "10",
day = "1",
doi = "10.1109/LRA.2018.2853651",
language = "English",
volume = "3",
pages = "3481--3488",
journal = "IEEE Robotics and Automation Letters",
issn = "2377-3766",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Motion switching with sensory and instruction signals by designing dynamical systems using deep neural network

AU - Suzuki, Kanata

AU - Mori, Hiroki

AU - Ogata, Tetsuya

PY - 2018/10/1

Y1 - 2018/10/1

N2 - To ensure that a robot is able to accomplish an extensive range of tasks, it is necessary to achieve a flexible combination of multiple behaviors. This is because the design of task motions suited to each situation would become increasingly difficult as the number of situations and the types of tasks performed by them increase. To handle the switching and combination of multiple behaviors, we propose a method to design dynamical systems based on point attractors that accept (i) 'instruction signals' for instruction-driven switching. We incorporate the (ii) 'instruction phase' to form a point attractor and divide the target task into multiple subtasks. By forming an instruction phase that consists of point attractors, the model embeds a subtask in the form of trajectory dynamics that can be manipulated using sensory and instruction signals. Our model comprises two deep neural networks: A convolutional autoencoder and a multiple time-scale recurrent neural network. In this study, we apply the proposed method to manipulate soft materials. To evaluate our model, we design a cloth-folding task that consists of four subtasks and three patterns of instruction signals, which indicate the direction of motion. The results depict that the robot can perform the required task by combining subtasks based on sensory and instruction signals. And, our model determined the relations among these signals using its internal dynamics.

AB - To ensure that a robot is able to accomplish an extensive range of tasks, it is necessary to achieve a flexible combination of multiple behaviors. This is because the design of task motions suited to each situation would become increasingly difficult as the number of situations and the types of tasks performed by them increase. To handle the switching and combination of multiple behaviors, we propose a method to design dynamical systems based on point attractors that accept (i) 'instruction signals' for instruction-driven switching. We incorporate the (ii) 'instruction phase' to form a point attractor and divide the target task into multiple subtasks. By forming an instruction phase that consists of point attractors, the model embeds a subtask in the form of trajectory dynamics that can be manipulated using sensory and instruction signals. Our model comprises two deep neural networks: A convolutional autoencoder and a multiple time-scale recurrent neural network. In this study, we apply the proposed method to manipulate soft materials. To evaluate our model, we design a cloth-folding task that consists of four subtasks and three patterns of instruction signals, which indicate the direction of motion. The results depict that the robot can perform the required task by combining subtasks based on sensory and instruction signals. And, our model determined the relations among these signals using its internal dynamics.

KW - AI-based methods

KW - Deep learning in robotics and automation

KW - humanoid robots

UR - http://www.scopus.com/inward/record.url?scp=85063308194&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063308194&partnerID=8YFLogxK

U2 - 10.1109/LRA.2018.2853651

DO - 10.1109/LRA.2018.2853651

M3 - Article

VL - 3

SP - 3481

EP - 3488

JO - IEEE Robotics and Automation Letters

JF - IEEE Robotics and Automation Letters

SN - 2377-3766

IS - 4

M1 - 8405582

ER -