- 著者
-
佐藤 大祐
美原 義行
佐藤 吉秀
田中 悠介
宮本 勝
佐久間 聡
- 雑誌
- 情報処理学会論文誌コンシューマ・デバイス&システム(CDS)
- 巻号頁・発行日
- vol.8, no.1, pp.1-10, 2018-01-30
本研究では,イベント会場における混雑度を即時的に把握・可視化することで,イベント運営者に混雑リスクの注意喚起をすることを目指す.イベント会場内の混雑度の取得のため,会場内にBLE(Bluetooth Low Energy)ビーコンを多数設置し,来場ユーザのスマートフォンアプリ(以下,アプリ)で取得したBLEビーコン電波情報をサーバで収集することで会場内の群集密度を計測する.1つのBLEビーコンがカバーする範囲内の人数として,そのBLEビーコンの電波を最も強く受信したユーザの数を数えることで求め,この人数をカバー範囲の面積で割ることで群集密度を求める.BLEビーコンを設置できない箇所も存在するため,空間内挿によりBLEビーコン地点間の混雑度を求める.本システムを,2日間で約5万人が来場する大規模なイベントに対して適用した.表示デザインの視認性の観点から判断し,10m程度の間隔でビーコンを設置し,ビーコンの出力電波強度を最大の+4dBmに設定した.イベント期間中の混雑時間帯において,会場で電波受信環境を調査したところ,圏外となった地点が存在しなかったことを確認した.そして,イベント期間中,本システムはアプリからの全120万件のビーコン電波受信ログのアップロードに対して,エラー率0で処理を完了させた.最終的な群集密度表示については,イベント運営者に実態と混雑度マップを比較いただき,実態と差異がなかったという評価をいただくことができた.In this research, we aim to alert the congestion risk to event operators by instantly grasping and visualizing the congestion degree at the event site. In order to acquire the congestion degree in the event site, we installed a lot of BLE (Bluetooth Low Energy) beacons and collected the BLE beacon information acquired by the visitor's smartphone application by the server. The server measures the congestion degree in the hall. Place the BLE beacon in the venue so that radio waves from at least one BLE beacon can be received at any point. There are no omissions in the number of installed users. In this research, the crowd density to be used as an indicator of congestion degree is obtained for correspondence of congestion risk. The number of people within the range covered by the BLE beacon is obtained by counting the number of users who received the radio waves of the BLE beacon most strongly and the crowd density is obtained by dividing this number by the area of a coverage area. Since there are places where BLE beacons can not be installed, the degree of congestion between BLE beacon points is obtained by spatial interpolation. In order to improve scalability, processing for finding the BLE beacon that received the strongest radio waves was cut out to the application side. In 2 days visitors offered this service with a large scale event of about 50,000 people. Regarding crowd density indication, we asked event operators to compare actual conditions and congestion degree maps, and received an evaluation that there was no difference with the actual situation. For the architecture that shared functions on the application side and the server side, processing was completed with an error rate of 0 with respect to the upload of all 1,200,000 beacon radio reception logs from the application. Furthermore, the effectiveness of being able to grasp the congestion degree in a bird's eye view from the operator was evaluated.