Case Study: Tenaris achieves rapid energy-demand forecasting and manufacturing efficiency with Cloudera

Tenaris predicting electrical energy demand with machine learning

Tenaris, a multinational maker of seamless steel pipes for the oil industry, faced limits from legacy systems that relied on small data samples and made it slow and difficult to combine diverse data sources or build predictive models. As analytical needs grew—requiring correlation of target variables with hundreds of features—Tenaris needed a modern data platform to scale analytics and accelerate model development.

By deploying Cloudera Enterprise, Tenaris ingests time-series data from thousands of sensors across five plants (capturing over 10,000 measurements per second) via Apache Flume and Sqoop, and processes it with Apache Spark, Impala and Tableau to train machine-learning models for energy prediction and production optimization. The solution reduced data‑science turnaround from weeks to days, improved product quality and energy efficiency, lowered costs, and enabled near‑real‑time decision making.

Tenaris

Vincenzo Manzoni

Head, Data Science