A preliminary study on a groping framework without external sensors to recognize near-environmental situation for risk-tolerance disaster response robots

Kui Chen, Mitsuhiro Kamezaki, Takahiro Katano, Taisei Kaneko, Kohga Azuma, Yusuke Uehara, Tatsuzo Ishida, Masatoshi Seki, Ken Ichiryu, Shigeki Sugano

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper proposes a basic near-environmental recognition framework based on groping for risk-tolerance disaster response robot (DRR). In extreme disaster sites, including high radiation and heavy smog, external sensors such as cameras and laser range finders do not work properly, and such sensors may be broken in accidents in the tasks. It is hoped that DRRs can continue to perform tasks, even if the external sensors cannot work, and at least, they can safely evacuate from the site. In this preliminary study, for recognizing near environments without using external sensors, we proposed a groping method. In this method, a robot actively touches the environment using arms or other movable parts, records the contact information, and then reconstructs a three-dimensional local map around the robot by the detected information, e.g., robot arm's position and reactive force. The proposed groping system can recognize the existence of three situations, such as an object, step, and pit, and those geometry, by exploring the designated space using arms. The groping strategy was designed considering both robot specification, time limitation, and required resolution. Experiments were performed using four-arm and four-crawler robot OCTOPUS. The results indicate that the proposed framework could recognize step, pit, and object, and calculate the position and size of the object, and confirm that the robot successfully removed the object on the basis of groped data.

Original languageEnglish
Title of host publicationSSRR 2017 - 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages181-186
Number of pages6
ISBN (Electronic)9781538639221
DOIs
Publication statusPublished - 2017 Oct 26
Event15th IEEE International Symposium on Safety, Security and Rescue Robotics, SSRR 2017 - Shanghai, China
Duration: 2017 Oct 112017 Oct 13

Other

Other15th IEEE International Symposium on Safety, Security and Rescue Robotics, SSRR 2017
CountryChina
CityShanghai
Period17/10/1117/10/13

Fingerprint

Disaster
robot
Disasters
tolerance
Tolerance
disaster
Robot
Robots
Sensor
Sensors
Laser Range Finder
Range finders
Framework
Accidents
accident
Extremes
Continue
Camera
Cameras
Radiation

ASJC Scopus subject areas

  • Artificial Intelligence
  • Safety, Risk, Reliability and Quality
  • Control and Optimization
  • Safety Research

Cite this

Chen, K., Kamezaki, M., Katano, T., Kaneko, T., Azuma, K., Uehara, Y., ... Sugano, S. (2017). A preliminary study on a groping framework without external sensors to recognize near-environmental situation for risk-tolerance disaster response robots. In SSRR 2017 - 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference (pp. 181-186). [8088161] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SSRR.2017.8088161

A preliminary study on a groping framework without external sensors to recognize near-environmental situation for risk-tolerance disaster response robots. / Chen, Kui; Kamezaki, Mitsuhiro; Katano, Takahiro; Kaneko, Taisei; Azuma, Kohga; Uehara, Yusuke; Ishida, Tatsuzo; Seki, Masatoshi; Ichiryu, Ken; Sugano, Shigeki.

SSRR 2017 - 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference. Institute of Electrical and Electronics Engineers Inc., 2017. p. 181-186 8088161.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Chen, K, Kamezaki, M, Katano, T, Kaneko, T, Azuma, K, Uehara, Y, Ishida, T, Seki, M, Ichiryu, K & Sugano, S 2017, A preliminary study on a groping framework without external sensors to recognize near-environmental situation for risk-tolerance disaster response robots. in SSRR 2017 - 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference., 8088161, Institute of Electrical and Electronics Engineers Inc., pp. 181-186, 15th IEEE International Symposium on Safety, Security and Rescue Robotics, SSRR 2017, Shanghai, China, 17/10/11. https://doi.org/10.1109/SSRR.2017.8088161
Chen K, Kamezaki M, Katano T, Kaneko T, Azuma K, Uehara Y et al. A preliminary study on a groping framework without external sensors to recognize near-environmental situation for risk-tolerance disaster response robots. In SSRR 2017 - 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference. Institute of Electrical and Electronics Engineers Inc. 2017. p. 181-186. 8088161 https://doi.org/10.1109/SSRR.2017.8088161
Chen, Kui ; Kamezaki, Mitsuhiro ; Katano, Takahiro ; Kaneko, Taisei ; Azuma, Kohga ; Uehara, Yusuke ; Ishida, Tatsuzo ; Seki, Masatoshi ; Ichiryu, Ken ; Sugano, Shigeki. / A preliminary study on a groping framework without external sensors to recognize near-environmental situation for risk-tolerance disaster response robots. SSRR 2017 - 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 181-186
@inproceedings{80578bfe7b3c47bd8420cd19971747bc,
title = "A preliminary study on a groping framework without external sensors to recognize near-environmental situation for risk-tolerance disaster response robots",
abstract = "This paper proposes a basic near-environmental recognition framework based on groping for risk-tolerance disaster response robot (DRR). In extreme disaster sites, including high radiation and heavy smog, external sensors such as cameras and laser range finders do not work properly, and such sensors may be broken in accidents in the tasks. It is hoped that DRRs can continue to perform tasks, even if the external sensors cannot work, and at least, they can safely evacuate from the site. In this preliminary study, for recognizing near environments without using external sensors, we proposed a groping method. In this method, a robot actively touches the environment using arms or other movable parts, records the contact information, and then reconstructs a three-dimensional local map around the robot by the detected information, e.g., robot arm's position and reactive force. The proposed groping system can recognize the existence of three situations, such as an object, step, and pit, and those geometry, by exploring the designated space using arms. The groping strategy was designed considering both robot specification, time limitation, and required resolution. Experiments were performed using four-arm and four-crawler robot OCTOPUS. The results indicate that the proposed framework could recognize step, pit, and object, and calculate the position and size of the object, and confirm that the robot successfully removed the object on the basis of groped data.",
author = "Kui Chen and Mitsuhiro Kamezaki and Takahiro Katano and Taisei Kaneko and Kohga Azuma and Yusuke Uehara and Tatsuzo Ishida and Masatoshi Seki and Ken Ichiryu and Shigeki Sugano",
year = "2017",
month = "10",
day = "26",
doi = "10.1109/SSRR.2017.8088161",
language = "English",
pages = "181--186",
booktitle = "SSRR 2017 - 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - A preliminary study on a groping framework without external sensors to recognize near-environmental situation for risk-tolerance disaster response robots

AU - Chen, Kui

AU - Kamezaki, Mitsuhiro

AU - Katano, Takahiro

AU - Kaneko, Taisei

AU - Azuma, Kohga

AU - Uehara, Yusuke

AU - Ishida, Tatsuzo

AU - Seki, Masatoshi

AU - Ichiryu, Ken

AU - Sugano, Shigeki

PY - 2017/10/26

Y1 - 2017/10/26

N2 - This paper proposes a basic near-environmental recognition framework based on groping for risk-tolerance disaster response robot (DRR). In extreme disaster sites, including high radiation and heavy smog, external sensors such as cameras and laser range finders do not work properly, and such sensors may be broken in accidents in the tasks. It is hoped that DRRs can continue to perform tasks, even if the external sensors cannot work, and at least, they can safely evacuate from the site. In this preliminary study, for recognizing near environments without using external sensors, we proposed a groping method. In this method, a robot actively touches the environment using arms or other movable parts, records the contact information, and then reconstructs a three-dimensional local map around the robot by the detected information, e.g., robot arm's position and reactive force. The proposed groping system can recognize the existence of three situations, such as an object, step, and pit, and those geometry, by exploring the designated space using arms. The groping strategy was designed considering both robot specification, time limitation, and required resolution. Experiments were performed using four-arm and four-crawler robot OCTOPUS. The results indicate that the proposed framework could recognize step, pit, and object, and calculate the position and size of the object, and confirm that the robot successfully removed the object on the basis of groped data.

AB - This paper proposes a basic near-environmental recognition framework based on groping for risk-tolerance disaster response robot (DRR). In extreme disaster sites, including high radiation and heavy smog, external sensors such as cameras and laser range finders do not work properly, and such sensors may be broken in accidents in the tasks. It is hoped that DRRs can continue to perform tasks, even if the external sensors cannot work, and at least, they can safely evacuate from the site. In this preliminary study, for recognizing near environments without using external sensors, we proposed a groping method. In this method, a robot actively touches the environment using arms or other movable parts, records the contact information, and then reconstructs a three-dimensional local map around the robot by the detected information, e.g., robot arm's position and reactive force. The proposed groping system can recognize the existence of three situations, such as an object, step, and pit, and those geometry, by exploring the designated space using arms. The groping strategy was designed considering both robot specification, time limitation, and required resolution. Experiments were performed using four-arm and four-crawler robot OCTOPUS. The results indicate that the proposed framework could recognize step, pit, and object, and calculate the position and size of the object, and confirm that the robot successfully removed the object on the basis of groped data.

UR - http://www.scopus.com/inward/record.url?scp=85040254940&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85040254940&partnerID=8YFLogxK

U2 - 10.1109/SSRR.2017.8088161

DO - 10.1109/SSRR.2017.8088161

M3 - Conference contribution

AN - SCOPUS:85040254940

SP - 181

EP - 186

BT - SSRR 2017 - 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference

PB - Institute of Electrical and Electronics Engineers Inc.

ER -