Using existing raw historical data, ML-algorithms will be developed and trained with tailored performance metrics. These algorithms will be implemented and evaluated for efficiency and applicability. The models will predict pump failures, considering factors such as pump locations, to optimize maintenance costs by servicing multiple pumps in a single visit. Our goal is to reduce maintenance costs, improve pump efficiency, and prevent pump failures by automatically alerting utilities of maintenance needs.
Grant source:
VUDP-foreningen (Vandsektorens forening til forbedring af vandsektorens effektivitet og kvalitet)
Granted amount:
805.355 mio DKK
Project start:
01/09/2025
