成果報告書詳細
管理番号20110000001177
タイトル*平成22年度中間年報 次世代ロボット知能化技術開発プロジェクト 移動知能(サービス産業分野)の開発 移動知能(サービス産業分野)の研究開発
公開日2011/10/12
報告書年度2010 - 2010
委託先名富士通株式会社 国立大学法人豊橋技術科学大学 株式会社セック 国立大学法人東京大学 国立大学法人奈良先端科学技術大学院大学 国立大学法人大阪大学 学校法人東京理科大学
プロジェクト番号P08013
部署名機械システム部
和文要約和文要約等以下本編抜粋:1. 研究開発の内容及び成果等 少子高齢化、労働力不足等の社会問題が顕在化する中、商業施設・交通施設・オフィス等時間的・空間的に変動し、人・障害物が混在する環境において、安全かつ適切な速度で移動し、案内・誘導・作業等の各種サービスを提供するロボットが、今後の高齢者の社会生活の質的向上及びサービス産業の発展に寄与する等大きな期待が寄せられている。 しかしながら、現状ではロボットが実用化できる用途や使用条件は限定的である。現在の自律移動ロボットは、誘導ガイド・ランドマーク・反射板等によって位置を同定し、予め入力した経路地図により移動しながら仕事を行う等、その活用範囲は限定的なものとなっている。
英文要約Title: Intelligent RT Software Project (FY2008-FY2011) FY2010 Annual Report
Fujitsu Limited: (1) We developed landmark map generation RTCs and landmark self-localization RTCs. We also developed obstacle detection and tracking RTCs that run on the image recognition module. (2) We developed an autonomous locomotion system comprised of our landmark self-localization RTCs, path planning RTC and so on. We verified the performance of our system and the reusability of our RTCs. Toyohashi University of Technology: We developed the following: (1) a SLAM and a localization method which use both appearance and range information, (2) observation planning RTCs, (3) person tracking methods which use overlapping silhouette templates and local features, (4) a motion planner based on 3D time-space inference, (5) a method of estimating object pose using a 3D range sensor. Systems Engineering Consultants Co.,LTD.: We are developing a middleware, which is the application framework and the test environment for Intelligent Mobile Robots. This middleware supports the latest version of OpenRTM-aist. And it provides a function to manage shared data among RT-Components and a framework for testing RT-Component easily. This middleware is useful and powerful for efficient construction of Intelligent Mobile Robots. Tokyo University: We developed (1)an algorithm to detect 3D visual odometry of the robot, (2)an algorithm to acquire image landmark, (3)an algorithm to detect ego-motion and moving objects' motions, (4)a guiding system for verifying the integrity of the developed RTC, (5)motion generation environment for HIRO-NX, (6)an evaluating system for SDL, (7)a program to find a box and grasp it by HIRO-NX. Nara Institute of Science and Technology: The basic modules for position estimation were developed using OpenRTM-aist-1.0.0 and integrated systems for mobile and manipulation were developed. The modules were integrated into the service robot "enon" and the reference hardware. We have started to develop a dual arm manipulation system that handles commodities and foods. This system will be used to evaluate modularity and reusability of the intelligent modules. Tokyo University of Science: We continuously made progress in the development of the person detecting and the person following modules for interpersonal movement. We developed a robust detection function under varying illumination based on H-S histograms. We implemented the developed method in the Reference Hardware Platform to confirm the validity of the proposed method. Osaka University: New pose estimation algorithm and its components have been developed to improve computation speed. In addition, component interface was fitted to the standard format, which was decided by project. All components, which were developed on OpenRTM-aist-0.4.2, were converted to OpenRTM-aist-1.0.0. As the components' reusability was improved, component layer structure was proposed and evaluated on Enon.
ダウンロード成果報告書データベース(ユーザ登録必須)から、ダウンロードしてください。

▲トップに戻る