
Electricity fraud has become a growing challenge for distribution companies due to its economic impact, public safety risks and negative effects on supply quality. To address this, Ariadna offers a solution based on a meter tampering detection algorithm that combines advanced analytics, AI and power-sector expertise to identify fraud automatically. The tool already operates across more than 5 million smart meters and achieves over a 95% reliability, helping to detect thousands of real cases.
At Ariadna, we are dedicated to researching and developing the necessary tools to enhance the management of electrical distribution networks.
Electrical networks are currently undergoing a major transformation. Until recently, distribution networks were largely static, consisting mainly of cables, transformers, and steel. However, about a decade ago, recognizing the need for greater information and control, various measurement devices began to be integrated into the networks, such as smart meters and feeder supervisors, enabling the collection of vast amounts of network data.
We focus on capturing this data and, through algorithmic tools and artificial intelligence, we design the electrical network of the future. We analyze the current state of the network, assess the impact of emerging technologies, and evaluate how the widespread adoption of electric vehicles and renewable energy sources will affect the existing infrastructure. Our goal is to identify the medium-term needs of the network and facilitate its evolution.
Detection of non-technical losses and fraud in LV networks (shunts and illegal connections).
One of the most pressing challenges today is the detection of electricity fraud. This issue has gained significant relevance in Spain, where fraudulent activities have increased sharply in recent years, particularly with the emergence of criminal organizations involved in illicit marijuana cultivation. As a result, electricity fraud has become not only a major concern for distribution companies but also a broader societal problem.
The most critical consequences of electricity fraud include:
- Economic impact: Energy losses translate into financial costs that are not only borne by distribution companies but are ultimately passed on to society, as these costs are shared among all consumers.
- Public safety: Illegal tampering with distribution networks often leads to accidents, posing serious risks to public health and safety.
- Supply quality: Fraudulent activities can compromise the stability and reliability of the power supply, affecting all consumers connected to the network.
A particularly common type of fraud involves meter tampering, an illegal practice in which users manipulate their electricity meters to record lower consumption than their actual usage, ultimately reducing their electricity bills
At Ariadna, we have developed a Meter Tampering Detection Algorithm designed to address this challenge. The key objectives of this algorithm are:
- Reducing energy losses, thereby lowering costs for society as a whole.
- Enhancing efficiency within distribution networks to facilitate the energy transition.
The main challenges we have faced in this project include:
- Developing the solution with a multidisciplinary team, comprising telecommunications experts, data scientists, and professionals from the electrical sector.
- Determining the optimal maturity level of the product before advancing to the next stage of development.
- Successfully integrating the solution into a real-world operational environment.

Reliable network topology for energy balancing.
Through collaboration with leading electricity distribution companies, this project has yielded significant breakthroughs, particularly in:
- Innovation: We have incorporated cutting-edge data management technologies, including artificial intelligence-based tools, deep learning techniques, and data mining methodologies. Additionally, cybersecurity measures have been implemented to ensure compliance with current regulations.
- Replicability and scalability: The algorithm has been designed for seamless integration into any distribution company, regardless of its location.
As a result, we have developed an algorithm that currently operates autonomously and automatically in more than 5 million smart meters, achieving an accuracy rate of over 95%.
Use Case: Loss Detection (Fraud/Shunt/Connections).
To date, we have successfully identified tens of thousands of manipulated meters, with field verification conducted by major electricity distribution companies. This validation underscores the effectiveness of our solution and highlights its contribution to combating electricity fraud.
