Sanning och konsekvens för svenska vattenledningar
This work kick-starts AI (artificial intelligence) for pipe management among Swedish water utilities. The current work with pipe management is mainly reactive; leaks are repaired after they are detected, sometimes with large costs if the leakage is extensive and critical. With this article, we want to focus on proactive pipe network management and discuss risk analysis in the field. A previously developed ANN model is used to estimate probability of leakage in water pipes. The model has been trained on leaks that have occurred over a ten-year period, and a comparison with leaks reported after the studied period shows that the ANN model succeeds in identifying groups of pipes with a higher leakage frequency. By combining the ANN model with a model for impact assessment, the most prioritised pipes, from a risk perspective, can be identified. While several water utilities have participated in the project Ordning i RörANN, only results based on data from Stockholm are presented here.