Active pressure managemente for a smart water network Machine Learning.


  • The Controller Peripheral does not need to know the pressure measurement at the Critical Point in order to set regulation correctly.
  • This means there is no longer any need for a communication channel between the two devices, and all related issues are therefore eliminated at source (e.g. to do with point to point communication, energy consumption).
  • The M.L. Algorithm protects against partial loss of data from one or both Peripherals
  • It is evolutionary and adapts daily to new data
  • It is analytical and recognises and filters faults in the data
  • It is verifiable and offers the operator all the tools to monitor the performance and degree of reliability of forecasting
  • It is controllable and allows the operator to define the level of autonomy of the system (control over sending of the Forecast Function to the Controller)

The New RTCP MACHINE LEARNING System consists of three subjects:

  • Peripheral Controller: active subject, forecasts the pressure at the Critical Point and applies it for Regulation.
  • Data Logger peripheral: passive subject
  • Centre-side Machine Learning algorithm: active subject. Learns the behaviour of the network and models the Controller peripheral.
rtcp machine learning Fast
Designed and assembled in Italy