成果報告書詳細
管理番号20160000000570
タイトル*平成27年度中間年報 次世代ロボット中核技術開発 革新的ロボット要素技術分野 知識の構造化によるロボットの知的行動の発現研究開発
公開日2017/2/15
報告書年度2015 - 2015
委託先名学校法人明治大学
プロジェクト番号P15009
部署名ロボット・AI部
和文要約
英文要約Title:Emergence of behavior in robotic systems by knowledge structuring (FY2015-FY2016) FY2015 Annual Report

The research objective is to put the robots can work in the human society properly by structuring knowledge with employing the results of AI research field. Followings are some of the progress reports in the fiscal year 2015.

I. Knowledge structuring and motion control
We've been working on the research to develop an intelligent control architecture which can activate a series of adequate motions at a right time based on the previous research on autonomous mobile robotic systems at Meiji university and on the recent progress of the field of artificial intelligence. In the year 2015, we have done some fundamental research and development of the control system, and tested its functions. The system can properly activate some motions, a robot which controlled by the system can move in the testing field successfully. In this experiment, the robot traversed indoor corridor, and ride on an elevator after press the button by its own. The robot can also wait the people have got out from the elevator before it get in. The series of motions have not programmed a priori, but are chosen by the system in every proper timings.

II. Automated map creation
Map information is essential for navigation with a goal. Our system could make the map by its own with its onboard sensors. Utilizing robot's precise localization and mapping capability, it's well known as a SLAM technology, three dimensional huge maps with very low distortion can successfully be created by integrating point-cloud data captured with 3D LiDAR. This technique can be available both indoor and outdoor, because it does not need GPS information. The map can also be involved semantic information.

III. Robotic hardware system development
The project will be planned to make robotic system can use the elevator or escalator which are the main infrastructures for pedestrians at a busy city. In the year 2015, sensor unit and mobility unit was developed.

IV. Cloud system
TIS corporation considered with Meiji university the architecture of cloud system that could be used for training of AI system of robots, simulations, and map sharing.
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