This paper presents the automation of the current leak detection operators experience and practice. A historical data driven statistical based approach is validated for real time leak detection in water distribution systems (WDS) yielding a color coded leak detection indicators (Likelihood LI, Severity SI, and Risk RI), which are integrated and visualized in an operator's user-friendly interface. This paper presents four signature scenarios using 3-year synthetic data from Eau de Paris monitored data. The data simulates leak scenarios in the WDS of the experimental demonstration site of Lille University Campus, France. The research includes: (1) assessment of the current state of practice; (2) development of an intelligent network control and on-site monitoring (INCOM) prototype system for early network leak detection; and (3) demonstration and assessment of the INCOM prototype system through off-line scenarios simulations. It supports the system operator(s) in "intelligent" real-time system monitoring and leak detection for preventive and proactive rather than reactive water distribution system management.