A modern approach for leak detection in water distribution systems

Water distribution is arguably the most important factor in modern times. The quality of service is nonetheless taken for granted while requiring regular maintenance and timely repairs. Therefore, there is an increased demand in efficient operation requiring modern solutions to age-old problems. Multiple theoretical and practical approaches exist for leak detection. In this paper, a model-based approach is extended with modern concepts from the field of Machine Learning for improving leak detection accuracy in water distribution systems.