Cyber-enhanced rescue canine

Kazunori Ohno, Ryunosuke Hamada, Tatsuya Hoshi, Hiroyuki Nishinoma, Shumpei Yamaguchi, Solvi Arnold, Kimitoshi Yamazaki, Takefumi Kikusui, Satoko Matsubara, Miho Nagasawa, Takatomi Kubo, Eri Nakahara, Yuki Maruno, Kazushi Ikeda, Toshitaka Yamakawa, Takeshi Tokuyama, Ayumi Shinohara, Ryo Yoshinaka, Diptarama Hendrian, Kaizaburo ChubachiSatoshi Kobayashi, Katsuhito Nakashima, Hiroaki Naganuma, Ryu Wakimoto, Shu Ishikawa, Tatsuki Miura, Satoshi Tadokoro

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

This chapter introduces cyber-enhanced rescue canines that digitally strengthen the capability of search and rescue (SAR) dogs using robotics technology. A SAR dog wears a cyber-enhanced rescue canine (CRC) suit equipped with sensors (Camera, IMUs, and GNSS). The activities of the SAR dog and its surrounding view and sound are measured by the sensors mounted on the CRC suit. The sensor data are used to visualize the viewing scene of the SAR dog, its trajectory, its behavior (walk, run, bark, among others), and its internal state via cloud services (Amazon Web Services (AWS), Google Maps, and camera server). The trajectory can be plotted on an aerial photograph captured by flying robots or disaster response robots. The visualization results can be confirmed in real time via the cloud servers on the tablet terminal located in the command headquarters and with the handler. We developed various types of CRC suits that can measure the activities of large- and medium-size SAR dogs through non-invasive sensors on the CRC suits, and we visualized the activities from the sensor data. In addition, a practical CRC suit was developed with a company and evaluated using actual SAR dogs certified by the Japan Rescue Dog Association (JRDA). Through the ImPACT Tough Robotics Challenge, tough sensing technologies for CRC suits are developed to visualize the activities of SAR dogs. The primary contributions of our research include the following six topics. (1) Lightweight CRC suits were developed and evaluated. (2) Objects left by victims were automatically found using images from a camera mounted on the CRC suits. A deep neural network was used to find suitable features for searching for objects left by victims. (3) The emotions (positive as well as negative) of SAR dogs were estimated from their heart rate variation, which was measured by CRC inner suits. (4) The behaviors of SAR dogs were estimated from an IMU sensor mounted on the CRC suit. (5) The visual SLAM and inertial navigation systems for SAR dogs were developed to estimate trajectory in non-GNSS environments. These emotions, movements, and trajectories are used to visualize the search activities of the SAR dogs. (6) The dog was trained to search an area by controlling the dog with the laser light sources mounted on the CRC suit.

Original languageEnglish
Title of host publicationSpringer Tracts in Advanced Robotics
PublisherSpringer-Verlag
Pages143-193
Number of pages51
DOIs
Publication statusPublished - 2019 Jan 1

Publication series

NameSpringer Tracts in Advanced Robotics
Volume128
ISSN (Print)1610-7438
ISSN (Electronic)1610-742X

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Artificial Intelligence

Fingerprint Dive into the research topics of 'Cyber-enhanced rescue canine'. Together they form a unique fingerprint.

  • Cite this

    Ohno, K., Hamada, R., Hoshi, T., Nishinoma, H., Yamaguchi, S., Arnold, S., Yamazaki, K., Kikusui, T., Matsubara, S., Nagasawa, M., Kubo, T., Nakahara, E., Maruno, Y., Ikeda, K., Yamakawa, T., Tokuyama, T., Shinohara, A., Yoshinaka, R., Hendrian, D., ... Tadokoro, S. (2019). Cyber-enhanced rescue canine. In Springer Tracts in Advanced Robotics (pp. 143-193). (Springer Tracts in Advanced Robotics; Vol. 128). Springer-Verlag. https://doi.org/10.1007/978-3-030-05321-5_4