• Leakage prediction in distribution networks
  • Optimal maintenance scheduling
  • No IoT or sensors needed
Betrieb & InstandhaltungDaten & AnalyseEnergieKünstliche IntelligenzWärmeWasser
Our cutting-edge machine learning framework can predict leakages in gas, water and electricity distribution networks. By knowing the quality of the network, Distribution System Operators can optimize the replacement policies and achieve significant savings in CAPEX/OPEX. No IoT or sensors needed. Model trained on historical data
We have developed a machie learning framework for assessment of the status of gas distribution pipelines and prediction of the future probability of leakage. The accurate prediction of gas leakages improves grid safety and reliability of supply and allows substantial savings on CAPEX (avoiding the replacement of pipelines which can stay in operations in the following years) and OPEX (prioritizing replacement of pipelines with a high probability of leakages, such that resources are not wasted on fixing gas leakages in the future).

The solution is based on an artificial intelligence model, coupled with advanced statistical modeling. The model is trained on the history of past leakages with additional information on the pipelines (e.g. material, location, length, protection and insulation type, etc) and augmented with external data (soil type, the pH level of the soil, tandmity of electrified railway lines, etc). The model learns patterns where the leakages occur and predict the probability of leakage for each segment of the pipeline in the future.

- No additional IoT devices and/or inspections are needed.
- Implementation in only 4 weeks

Unsere Mitarbeiterinnen und Mitarbeiter



186 00 Prague

Jetzt kostenfrei anmelden

Melden Sie sich jetzt kostenfrei auf KommunalDigital an, um von weiteren Features zu profitieren.