FIT2016 第15回情報科学技術フォーラム 開催日:2016年9月7日(水)~9日(金) 会場:富山大学キャンパス
井上千徳・矢口勇一(会津大)・佐藤 彰・菊池 稔(アクアクルー)・岡 隆一(会津大)
To observe and analysis sewage plant facilities, we propose a visualization technique which is clustering state and drawing state pattern on an image map. Aqua-crew and the University of Aizu developed a remote monitoring system for indicator board in sewage plants, and this system stored on cloud server which recognized lamp signals on indicator boards. Our proposal is to make cluster from this high-dimensional signals to labeled signals by cluster number, and plot on an image map for visualizing. From the experimental result,we can easily find stable and illegal pattern, also small cluster has "failure" lamp in high ratio.