Artificial Intelligence Based Framework for Industrial Asset Management Using Real, Time Series Data

This Research delivers a Self-Exploratory Deep Learning Hybrid Framework that uses real-time data generated by industrial sensors attached to an Industrial Asset to perform real-time predictions related to asset health. The Self-Exploratory Framework is designed to discover and explain hidden relationships between a predicted variable and its co-variates and can handle both univariate and multi-co-variate sensor data for future value prediction. The Framework is scalable and incorporates multiple models to derive the best prediction strategy. Its self-exploration ability continually monitors and adapts to changing machine behaviour, reduces manual intervention, provides deeper insights into asset health and optimises asset utilisation.