We apply machine learning models and intelligent visualization to improve efficiency and quality at Ercros’ plant.

Ercros is an industrial group with a long-standing history, diversified across various areas, including the Intermediate Chemicals Division. In this context, its Tortosa plant is a key center: the only one in Spain and the fourth worldwide in the production of pentaerythritol and dipentaerythritol, essential compounds for the manufacture of paints and lubricants.

With the aim of improving the efficiency and stability of its production process, Ercros has partnered with ThinkUPC to develop an advanced monitoring system based on machine learning techniques. The project includes the implementation of two models —a signal behavior autoencoder and a multivariable model based on PCA— as well as a custom application for real-time data visualization.

With this initiative, Ercros expects to achieve a significant reduction in process variations, improved product quality, and faster, better-informed decision-making. A positive impact is also expected in terms of cost reduction and resource optimization.

Find out how we are tackling this challenge by reading the full case study.