Case Study: Madrileña Red De Gas boosts fraud detection from 5% to 30% and reduces non-technical losses with Atos Codex

For Spanish utility company, Madrileña Red De Gas, intelligent analytics have dramatically boosted fraud detection rates and reduced non-technical losses

Madrileña Red De Gas, a Spanish regional gas distributor with about 5.6 thousand km of distribution pipe and some 850,000 customers, faced significant non-technical losses (customer theft) and lacked the historical records and intelligence to target inspections effectively. To tackle this, the company engaged Atos and its Atos Codex analytics capabilities to explore how data analytics and IoT could improve detection and make field inspections more productive.

Atos used Atos Codex frameworks, high-performance computing and iterative data‑science models to analyze usage data and prioritize suspect sites for inspection. The approach lifted fraud detection hit rates from around 5% pre-analytics to 20% after the first wave and to 30% after the second wave — all in less than 12 months — while also uncovering unexpected patterns (for example, higher fraud in detached houses), delivering measurable reductions in non-technical losses and ongoing improvements through continuous analytics by Atos.

Madrileña Red De Gas

Glen Lancastle

Chief Information Officer & Chief Operactions Officer