ロボット技術を利用した実世界の情報収集  [in Japanese] Robotics technologies for gathering information in the real world  [in Japanese]

Access this Article

Search this Article

Author(s)

Abstract

著者は,ロボット技術を利用した実世界の情報収集に関して研究を行っている.情報化が進む現代においては,変化する情報をすぐに積極的に収集し,解析することが欠かせない.実世界を動くことが出来るロボットは新たな情報収集の手段になる.著者の所属するグループでは,被災した建物内を探査するクローラロボット,老朽化した構造物の点検を行う飛行ロボット,嗅覚で被災者を探査する災害救助犬のデジタル化,屋外を自動で走行する無人搬送車などを開発している.災害地の探査,老朽インフラの点検,公道の自動運転に関する社会的なニーズは,東日本大震災後,また 2020 年の東京オリンピックを前に高まっている.その結果,産学官が参加した大きなプロジェクトも起こりつつある.瓦礫や橋梁などの狭い場所を探査するロボットは,搭載出来る機器の大きさや重さに制限があり,小さくて軽いセンサしか搭載できない.この結果,過酷な条件で質の悪いデータが大量に記録される.このような質の悪いセンサデータから,人や傷などの対象を認識し,その位置を特定するために形状や軌跡を復元することはチャレンジングな研究テーマである.コンピュータビジョンの研究者の協力により,質の悪いデータでも従来と同等の形状や軌跡の復元が行えることで,著者らが開発しているロボットが社会で利用されことにもつながる.ロボットが集める質の悪いデータが,社会を動かす石油にもなりえる.本稿では,実世界で情報収集を行うロボットの開発と,そのロボットから得られるセンサデータについて紹介する.The author has studied about robotics technologies for gathering information in the real world. In the present age when computerization advances, we need to gather information of the real world and to find its change by analyzing them, immediately. The robot, which can explore the real world, becomes the mean of new intelligence. Our group developed tracked vehicles for surveying damaged buildings, flying robots for inspecting the surface of old bridges, digitally empowered search and rescue (SAR) dogs for finding victims, and unmanned autonomous carriers for carrying objects in outdoor environment. Social needs about the exploration of the disaster areas, inspection of the old infrastructure, and autonomous driving in public roads increases after East Japan great earthquake and before Tokyo Olympic in 2020. National projects of industry-university-government cooperation have started in Japan. The rescue robots and inspection robots does not have enough capacity for carrying the sensor devices. These robots can be equipped with only small-size and lightweight sensors, which produce huge and low quality data. The object recognition and map reconstruction from low quality sensor data will be challenging research topics for researchers in robotics and computer vision. Cooperation of robotics and computer vision enables us to recognize them and to reconstruct map from low quality data. These results promote the progress of robotic applications. Low quality data becomes new oil, which innovates the robotic application. This presentation introduces our robotics technologies that gather information in the real world, and also show its sensor data.

The author has studied about robotics technologies for gathering information in the real world. In the present age when computerization advances, we need to gather information of the real world and to find its change by analyzing them, immediately. The robot, which can explore the real world, becomes the mean of new intelligence. Our group developed tracked vehicles for surveying damaged buildings, flying robots for inspecting the surface of old bridges, digitally empowered search and rescue (SAR) dogs for finding victims, and unmanned autonomous carriers for carrying objects in outdoor environment. Social needs about the exploration of the disaster areas, inspection of the old infrastructure, and autonomous driving in public roads increases after East Japan great earthquake and before Tokyo Olympic in 2020. National projects of industry-university-government cooperation have started in Japan. The rescue robots and inspection robots does not have enough capacity for carrying the sensor devices. These robots can be equipped with only small-size and lightweight sensors, which produce huge and low quality data. The object recognition and map reconstruction from low quality sensor data will be challenging research topics for researchers in robotics and computer vision. Cooperation of robotics and computer vision enables us to recognize them and to reconstruct map from low quality data. These results promote the progress of robotic applications. Low quality data becomes new oil, which innovates the robotic application. This presentation introduces our robotics technologies that gather information in the real world, and also show its sensor data.

Journal

  • IPSJ SIG Notes. CVIM

    IPSJ SIG Notes. CVIM 2015-CVIM-196(16), 1-1, 2015-02-27

    Information Processing Society of Japan (IPSJ)

Codes

  • NII Article ID (NAID)
    110009882601
  • NII NACSIS-CAT ID (NCID)
    AA11131797
  • Text Lang
    JPN
  • Article Type
    Technical Report
  • ISSN
    09196072
  • Data Source
    NII-ELS  IPSJ 
Page Top