If you prefer to always view the website in English, please click here.
Images of different types of fraud detected in the distribution network.
- The company's distribution subsidiary, e-distribución, applies machine learning and deep learning techniques to uncover cases of electrical fraud.
- In the last year, it has detected energy theft in one in every two inspections, twice the number in 2017, when it began to introduce artificial intelligence to combat fraud.
- An Endesa team made up of mathematicians, computer scientists and engineers is leading the use of AI to combat fraud.
In just three years, due to the introduction of artificial intelligence the success rate of inspections against fraud carried out by Endesa has shot up.
e-distribution, Endesa's distribution subsidiary, carried out more than 600,000 inspections in the 2018-2020 period to detect electrical fraud. And in the last year it has managed to uncover energy theft in one out of every two inspections, a percentage that doubles that registered in 2017 when the company began to apply machine learning and deep learning techniques to detect non-technical energy losses.
The digitisation of the network, the deployment of sensors and the implementation of smart meters means that more and more information is being obtained on the operation of measurement equipment and the medium and low voltage network. Data analysis enables deviations and anomalous behaviour to be detected so that inspections are directed more efficiently and there is an increase in the percentage of fraud detected.
Always adhering to regulations, e-distribution works with a significant volume of data that is put into a data lake on which advanced programming languages are used to create models that allow fraud to be detected.
The application of artificial intelligence on data and the systematic improvement of these predictive models is enabling Endesa to efficiently detect all types of fraud, both in supply lines with a valid contract, and in supply lines without a contract, from traditional types of fraud (bridges, double connections, direct couplings, etc.), even the most sophisticated (boards with printed circuits inserted in the meters themselves), detecting a large amount of fraud and with a precision that would have been difficult to imagine just a few years ago.
The Machine Learning unit is made up of mathematicians, computer scientists and engineers who apply artificial intelligence in the fight against electricity fraud at Endesa. They are experts in data science and big data. These new professional profiles have joined the Endesa staff in recent years and work together with field inspectors to develop and improve predictive models.
This line of work reflects the clear commitment of Endesa and the entire Enel Group to become a data driven company and to take advantage of the large volume of data it manages to improve operations and decision-making.
About Endesa
Endesa is the leading electricity company in Spain and the second largest in Portugal. It is also the second largest gas operator in the Spanish market. It is an integrated business operation that encompasses everything from generation to marketing, and through Endesa X it provides added value services aimed at decarbonising the energy used in homes, companies, industries and government agencies. Endesa is firmly committed to the United Nations SDGs and as a result decisively promotes the development of renewable energies through Enel Green Power Spain, the electrification of the economy and Corporate Social Responsibility. We also work in the latter area through the Endesa Foundation. We have around 10,000 employees. Endesa is part of Enel, the largest electricity group in Europe.