The historical data of wastewater treatment plants hides the operating experience of the employees. This data-based operational and empirical knowledge needs to be leveraged and reused.
nerou's software analyzes this historical data from system operation and uses it to train the machine learning algorithms (self-learning algorithms). If the software detects a change in the live values, an algorithm searches for the most similar situation in the historical data.
Once this situation is found, the algorithm examines which value arose from this situation. Another algorithm developed by nerou then determines the best possible response for the current situation and gives a recommendation or a control impulse as to how the sewage treatment plant should be controlled.
The three areas of data analysis, biological and process engineering expertise and software development come together to create these decision-making aids.