The PredictIA project, developed by Moeve in collaboration with the University of Malaga, is co-financed by the Center for Industrial Technological Development (CDTI), under file IDI-20250628, and by the Technology Corporation of Andalusia (CTA), reference 25/1180.
Its overall objective is to progress in the optimization of complex industrial processes, addressing both the blending of diesel A and the Hydrocracking unit of the La Rábida Energy Park. In the case of blending, the aim is to optimize the quality of various diesel mixtures without compromising the specification of critical properties such as viscosity. Although the project encompasses both areas, it is in the Hydrocracking unit where the Innovation Center for Energy Transition plays a key role, thanks to its deep understanding of the operational and catalytic behavior of the process.
With this in mind, the work is focused on the development of an advanced decision-support tool for that unit. The solution integrates Artificial Intelligence architectures with models based on fundamental scientific principles, allowing the prediction of catalyst deactivation and other key variables for operational planning. In this way, PredictIA complements and enriches existing systems, providing a predictive vision of high value for daily operation.
The model is fed with large volumes of historical data, coming from both industrial sensors and laboratory analysis, which allows for real-time estimation of the unit's condition. Thanks to this, PredictIA is configured as the foundation of an intelligent decision-support system, capable of proposing operational adjustments that maximize economic benefit, reduce consumption, and help minimize the environmental footprint.
