Optimizing network replacement with AI

Abstract

With a network of 1.4 million km Enedis spends several hundred million Euros every year to maintain and renew the existing network by replacing parts of it.To be able to replace the network parts which are the most likely to be source of future faults is crucial for good capital expenditure (CapEx) management as it lowers operational expenditure caused by maintenance or incentive regulation rules.For the past three years Enedis has been using Artificial Intelligence (AI) to optimally choose which parts of the network have to be renewed.In this paper we detail the method, which was initially successfully developed for underground Low Voltage (LV) networks, and point out how it has been extended to underground Medium Voltage (MV) networks and overhead networks (LV and MV).(remark: this a non-defining relative clause, so you have to put a comma and can`t omit the relative pronoun)We also describe how big data technologies have allowed us to leverage on the work done to target network replacement in order to initiate a transition towards predictive maintenance.