RTCP ML

RTCP-ML (Real-Time Critical Point System – Machine Learning) is an intelligent system for the automatic and dynamic regulation of pressure in water distribution districts. Based on a predictive algorithm, it learns the hydraulic behavior of the network and estimates the pressure at the Critical Point (CP) using only local data (inlet pressure, outlet pressure, and flow), without the need for direct measurement. In the event of communication interruptions between devices, the system automatically switches from real-time data to estimated values, ensuring continuous operation. An intuitive interface allows real-time monitoring and the automatic sending of alerts via SMS or email in case of anomalies.

The artificial intelligence of RTCP-ML allows to anticipate and proactively address critical issues, promoting increasingly resilient, efficient, and smart water network management.

• Operational simplicity: the regulator peripheral does not need to receive the measured value from the critical point, simplifying the system architecture.
• No direct communication required: removing the point-to-point channel between peripherals reduces complexity, energy consumption, and communication-related risks.
• Robust and adaptive algorithm: the Machine Learning model is resilient to missing or incomplete data and continuously updates as the network evolves.
• Analytical and transparent: it detects and filters any data anomalies and provides advanced tools to monitor performance and forecast quality.
• Flexible control: the operator can define the system’s level of autonomy, deciding when and how to apply forecasts to regulation.
• System composed of a regulator peripheral, which forecasts the pressure at the critical point (CP) and applies the calculated values to regulate the PRV; a critical point peripheral, which collects data and sends it to the central system; and a central Machine Learning algorithm, which learns from historical and real-time data, models the network’s behavior, and supports the regulator peripheral.