本文へジャンプ

成果報告書詳細
管理番号20190000000121
タイトル*2018年度中間年報 次世代人工知能・ロボットの中核となるインテグレート技術開発 人工知能技術の社会実装に向けた研究開発・実証 人工知能技術を用いた安心・安全・快適で効率的なオンデマンド乗合型交通の実現
公開日2019/4/16
報告書年度2018 - 2018
委託先名株式会社未来シェア 国立研究開発法人産業技術総合研究所 株式会社NTTドコモ
プロジェクト番号P18002
部署名ロボット・AI部
和文要約
英文要約Title: Demand-Responsive Sharing Transportation System by Artificial Intelligence (FY2018-FY2019) FY2018 Annual Report

In this fiscal year, we carried out the following four researches and developments:

(1) We established systems and platforms for actual services of the Smart Access Vehicle Service (SAVS), and developed analysis methods to adjust and improve policies of vehicle assignments and operations. Because the experimental service described in (4) was expected the largest scale of SAVS operation, we designed the systems and platforms reliable enough for various changes and types of demands. The developed methods are tested and tuned in agile way by using real data taken by the experimental service

(2) We constructed frameworks to evaluate the policies of vehicle assignments and operations proposed in (1) under various changes of demands. The framework enables exhaustive analysis of various frequencies of demands, and measures the performance of the operation by usability and effectivity of SAVS. We also established a method to factorize real OD data taken in the experiment of (4) into elements of demand distributions that reflect features of demands of categorized days/hours. These results will enable to evaluate various unseen situations of demands-occurrence.

(3) We built up a hypothesis about resource managements of SAVS for various operational conditions in order to maximize efficiency of the operation with keeping an acceptable service level for users. To make the experiment held in (4) as successful as possible, we conducted the hypothesis about the number of operated vehicles to provide enough satisfaction for users with minimized operational costs. The hypothesis was refined using the real data and experiences of the experiment toward the next experiments.

(4) We carried out an experimental service at Minato-Mirai and Kannai area during 10 weeks in October, November and December, 2018 with the collaboration of Yokohama-city. The size of the service area was 2km x 5km, and 10-15 vehicles were operated in each day. The service enabled transportation among entire regions. Finally, the service carried about 34,000 passengers, and got enough number of data that reflects various types of demand-frequencies, which were utilized to evaluate and refine the results of works in (1), (2) and (3).
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