Abstract
In order to cope with the decrease in the number of operation managers at wastewater treatment plants, it is important to develop and introduce AI technology that can reproduce the operation of skilled managers and contribute to transfer operation know-how. In this study, a questionnaire survey was conducted to understand the demand for AI technology and the operation items that AI technology can guide in wastewater treatment. The results indicated that AI technology could be utilized at wastewater treatment plants, in particular, to help operation managers decide the set values of airflow rate of reaction tanks and excess sludge withdrawal rate, and that there is a certain level of demand for AI technology from operation managers. In addition, we introduced AI technology to an actual wastewater treatment plant to support response decisions and operations, and conducted a demonstration experiment. The AI's inferences for 12 operational items were compared with the judgments of the skilled managers at the wastewater treatment plant, and the results showed that the judgments agreed in approximately 90%. This result indicates that AI technology can be used to help operation managers make decisions on the actual operation of wastewater treatment plants.
