| 田中 智基 | D, 中間発表 | ユビキタスコンピューティングシステム | 安本 慶一, 笠原 正治, 諏訪 博彦, 藤本 まなと, 松田 裕貴 |
|
title:Evacuation Shelter Selection Method Considering the Assembly of Family Members
abstract:In disaster situations, being apart from family members causes severe mental damages to evacuees. Particularly, the disaster weak, such as the elderly, disabled people and children can be under stress due to absence of their guardians. In this study, we reinvented our conventional shelter selection method for both minimizing evacuation time and gathering family members in remote locations. Assuming scenarios of disaster situations in Ikoma city, we conducted multi-agent simulations for the conventional method and the new method and compared total evacuation time and how many family households were able to gathered. As the result, we have confirmed the trade-off relationship between evacuation time and gathering family members as well as the difference of evacuation time by the settings of the new shelter selection method. language of the presentation: Japanese 発表題目: 災害時の世帯構成員の集合を考慮した避難所決定の検討 発表概要: *** 災害発生時において、家族が離れている状態は被災者の精神的負担の大きな要因となる。特に、高齢者や身障者、子供といった災害弱者にとってその負担の影響は著しいものである。本研究では、従来手法を発展させ、避難時間の最小化に加え、家族の集合を目的とした避難所決定手法を提案する。生駒市の地震発生時を想定したシナリオを元に、従来手法及び新たに世帯集合を目的とした避難所決定手法のシミュレーションを実行し、避難完了時間、世帯集合度の比較を行った。結果、世帯集合度と避難完了時間のトレードオフの関係、世帯避難時間の設定による避難完了時間の変化が確認された。 *** | |||
| 佐瀬 凌太 | M, 2回目発表 | ユビキタスコンピューティングシステム | 安本 慶一, 松本 健一, 諏訪 博彦, 藤本 まなと |
|
title: A Method for Predicting Vehicle Returns in a One-Way ECS Using Real-Time Movement Information
abstract: In recent years, interest in One-Way Electric Car Sharing services (ECS), which are sustainable mobility, has been growing. While One-Way ECS allows more flexible use by users, its usage pattern causes a vehicle uneven distribution problem. In this study, we propose a vehicle return prediction model for One-Way ECSs to solve this problem. In addition to the information on the user's starting location and time of use, the proposed method aims to continue to obtain more accurate prediction results by taking into account time-series changes in location and direction of movement. In an evaluation using actual usage data of a One-Way ECS, the proposed method achieves an accuracy of 0.93 for the prediction of stations to be returned, and an MAE of 42.3 min. and a MAPE of 47% for the prediction of return times. Furthermore, the results confirm the improvement of the prediction results over time. language of the presentation: Japanese 発表題目: リアルタイム移動情報を用いた乗り捨て型ECSにおける車両返却予測手法 発表概要: 近年，サステイナブルなモビリティである乗り捨て型電機自動車シェアリングサービス(ECS)への関心が高まっている．乗り捨て型ECSは，ユーザによりフレキシブルな利用を可能にする一方で，その利用形態から車両偏在問題が発生してしまう．本研究では，この問題の解決に向けた乗り捨て型ECSにおける車両返却予測モデルを提案する．提案手法では，ユーザの利用開始場所・時間の情報に加えて，時系列的に変化する位置情報・移動方向情報などを考慮することでより正確な予測結果を得続けることを目指している．提案手法は実際の乗り捨て型ECSの利用データを用いた評価において，返却されるステーションの予測では0.93という正解率，返却時間の予測では42.3minのMAE及び47%のMAPEを達成している．さらに，時系列的に予測結果が改善されることが確認できている． | |||
| 真弓 大輝 | M, 2回目発表 | ユビキタスコンピューティングシステム | 安本 慶一, 清川 清, 諏訪 博彦, 松田 裕貴 |
|
title: Aromug: Mug-type Olfactory Interface to Reduce Sugar Intake
abstract: Reducing sugar intake from daily beverages is essential for maintaining good health. However, for those who prefer sweetened beverages, it is mentally distressing to suddenly reduce their intake, and a solution is needed to help them gradually become accustomed to sugar-free beverages. In this paper, we propose the Aromug, a mug-type olfactory interface that amplifies the user's perception of sweetness by providing a sweet aroma along with the act of drinking. Through a user study, we investigated the influence of aroma information presented by the Aromug on the taste of iced coffee. The experimental results showed that the aroma of unsweetened coffee and chocolate could amplify the perceived sweetness of iced coffee compared to that of unflavored coffee. The results also suggested that there were differences in sweetness preferences depending on age and frequency of coffee drinking. language of the presentation: Japanese 発表題目: Aromug: 糖分摂取量低減を補助するマグカップ型嗅覚インターフェース 発表概要: 日々の飲料から摂取する糖分を減らすことは，健康維持のために欠かせない．しかし，甘い飲み物を好む人にとって，急に摂取量を減らすことは精神的に苦痛であるため，徐々に無糖の飲み物に慣れていくための解決策が必要とされている．本論文では，飲むという動作とともに甘い香りを提供することで，ユーザが感じる甘さを増幅させるマグカップ型嗅覚インターフェース「Aromug」を提案する．ユーザースタディを通じて，Aromugが提示する香り情報がアイスコーヒーの味に与える影響を調査した．実験の結果，無糖コーヒーとチョコレートの香りは，無香料の場合と比較して，知覚する甘さを増幅する可能性が見られた．また，年齢やコーヒーの飲む頻度によって甘さの嗜好性に違いがあることが示唆された． | |||
| DE LEON AGUILAR SERGIO | M, 2回目発表 | ユビキタスコンピューティングシステム | 安本 慶一, 向川 康博, 諏訪 博彦, 松田 裕貴 |
|
title: Designing a More Engaging Disaster Prevention Guideline through Mobile Technologies
abstract: Disaster Prevention Guidelines provide instructions to the citizens to improve their reactions before, during, and after disasters. Our previous work has unveiled the citizenry's tendency to not read or follow their recommendations. Also, they are regarded as hard to follow. Research in educational and industrial environments has proven Augmented Reality (AR) technologies, as effective in reducing errors and improving the learning experience. However, their designs are heavily dependent on the age and the context of the user, a specificity the general population lacks. We now present a novel interface design intended to increase citizen engagement and the effectiveness of Disaster Prevention Guidelines. AR technologies provide close-to-real navigation through the targets of the guidelines. In addition, a deep learning object detection model helps reduce the dependency of the user on their inexpert judgment in recognizing such targets of interest of the guidelines. Through this platform, our study takes the contributions of previous research in user interfaces that targets a specific public, and tries to apply such lessons to a broader category of the public. language of the presentation: English | |||