Customers: Main Power Networks of the Center Contractors: Mail.ru Group Product: Artificial intelligence (AI, Artificial intelligence, AI)Project date: 2020/09 - 2021/02
|
2021: Development of a complex to combat commercial loss of electricity
March 17, 2021 Mail.ru Digital technologies announced the development of a software complex based on artificial intelligence, which helped Rosseti Center more effectively combat commercial losses of electricity.
Every year, electric power plants suffer heavy losses due to illegal connections of legal entities to electricity grids, interference with the operation of metering devices and other malfunctions, abuses. Previously, the Rosseti Center field teams conducted more than 646,000 inspections annually in search of facts of unaccounted for electricity consumption. A month, company specialists could analyze no more than 30% of accounting points. At the same time, in 99% of cases, inspections did not reveal violations, since it is difficult for a person to take into account all the factors affecting the loss of electricity.
Rosseti Center identified the reduction of commercial losses as one of the priority tasks in the company's development strategy and turned to the PREDICT team from Mail.ru Group to develop a solution. To increase the fight against commercial losses, PREDICT specialists have created a software complex based on artificial intelligence and machine learning technologies.
The system analyzes its own data, for example, the status of the accounting point, the dynamics of electricity consumption, and information from open sources - information about the legal entity and others. This allows the software complex to identify non-trivial links between targets and real indicators at each metering point, calculate the probability and predict the amount of unaccounted for energy consumption.
If the probability of commercial loss at a particular metering point is determined to be high, the system automatically generates a task for the field team. After completing the task, the program analyzes the checks performed and uses data the ML models obtained for further training.
The project proved expediency and demonstrated the real economic effect of the use of artificial intelligence in the field of electric power. The solution developed by the PREDICT team allowed the company to analyze up to 100% of electricity consumption points every month and at the same time increase the efficiency of finding violations by 2.2 times. At the same time, the average cost of detecting one fact of unaccounted consumption decreased more than twice: from 36,600 to 16,900 rubles.
Artificial intelligence covers more and more areas. We are pleased that our decision allowed Rosseti Center to more effectively identify violations related to unaccounted for energy consumption by legal entities and reduce commercial losses. And, of course, we hope that this example will be an impulse for other energy industry companies that were thinking about using AI, "commented Roman Styatyugin, head of analytical services at PREDICT, Mail.ru Group. |
Artificial intelligence technologies have great potential in the context of strategic development and increasing the efficiency of processes at Rosseti Center, "said Vitaly Akulichev, deputy general director for digital transformation at Rosseti Center. - Integration of analytical solution is the first step in the series of digital transformation projects of our company. And we are confident that our technological cooperation with Mail.ru Group will be long-term and productive. |