2020 Fiscal Year Annual Research Report
Autonomous wastewater treatment: Monitoring and control of on-site wastewater treatment plants
| Project/Area Number |
20F20763
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
國吉 康夫 東京大学, 大学院情報理工学系研究科, 教授 (10333444)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SCHNEIDER MARIANE 東京大学, 情報理工学(系)研究科, 外国人特別研究員
|
|---|
| Project Period (FY) |
2020-11-13 – 2023-03-31
|
| Keywords | wastewater treatment / soft sensors / monitoring / control / partially observable / machine learning / hybrid models |
|---|
| Outline of Annual Research Achievements |
The first achievement is a detailed five page research statement for my research in the next 4 years. This allowed me to get shortlisted for a Tenure track position at the Swiss Federal Institute of aquatic research and to further refine my research questions. A new aspect added to the plan by this is that I plan to systematically model possible wear-and-tear effects of sensors in order to test with the model the influence of these effects on monitoring and control. This is more resource-efficient, allows extrapolation, and allows me to make the experiment more target oriented (to check the physical model and fill the gaps which cannot be answered by the model alone). I will test the control with the dissolved oxygen sensor, as I can program a simple model myself, which allows me to tweek all parameters to include the wear-and-tear effects. Second, a scientific article soon to be submitted on the effects of sensor inaccuracy on the performance of an entire system of on-site wastewater treatment plants. Third, participation in a competition on digitalization of water treatment in collaboration with other members of the Kuniyoshi laboratory, which allowed me to learn some of the methods in robotics and machine learning and establish a collaboration despite remote work. We made it to the final round of the competition. The results will be announced at the beginning of May.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
The project is advancing well, but not fully as planned. The experimental part is slightly behind schedule and will get even further behind schedule due to limiting the time in the laboratory (Covid-19 spread). However, by thinking about an alternative, a modelling approach came up, which is in my opinion not just better adapted to the current situation, but will also benefit my research, especially as I can benefit from the modelling experience in the AI research center at the University of Tokyo where I am affiliated with. The exchange with the group at the University of Tokyo is also going slower due to remote work. Therefore, I plan to get involved in a research project of one of the researchers of the group. On the other hand, the research question has refined to the desired level of detail and collaborations also with Kyoto University have been established and work very nicely. I furthermore, could contribute to a project on modelling behavior in sewers in order to monitor sewers, which is conducted at Kyoto University. So new content was produced, which is a success. ）
|
| Strategy for Future Research Activity |
Due to the additional task of modelling the wear-and-tear effects of sensors on control, the experiment will take place later than planned and in a more concentrated form (shorter), as it makes sense to first explore with a model the potential effects in order to better plan and conduct the experiment and come up with a more generalisable result.
|
