Department Urban Water Management

Mr. Rohr arrives at his workplace at the sewer utitily Cleanlake on Monday morning. He has to ensure the functioning of 20 rainwater tanks and pumping stations of 8 communities, to repair failures and plan operations. In the past, remote monitoring alone required 3 hours per week - an unpopular job that already led to bad mood at the beginning of the week. Today, the virtual employee RICHARD (RICH Analytical Rule-engine for operational and strategic Decision support), who is working day and night, has already done the analysis of the pumping cycles, the water levels and the power consumption. Overall, there is little wastewater flow, but this is not unusual for spring holidays. Mr. Rohr is reassured. However, RICHARD has also carried out the 3-monthly infiltration water analysis and points out that 20% more clean groundwater enters the network than permitted by the municipality Investmentlazy. He must discuss this with his supervisor. RICHARD also recommends medium-term maintenance of pump no. 23, but the diagrams do not yet show any need for action for Mr. Rohr.

Although process control systems are already used in many sewer networks to collect measurement data, these data is not evaluated. Due to low data competence, even simple analyses are perceived as laborious and are assigned to specialists, which consumes valuable resources. This prevents regular evaluations, which in turn reduces data quality.

In this 2-year project, appropriate building blocks are created using modern big-data methods and technologies, such as data lakes and AI, in order to effectively automate recurring tasks. The foundations for more extensive evaluations, such as key figures and forecasts, are created, which increases data competence and enables secondary uses.