The concept of the Digital Twin, defined as a virtual representation of a physical asset, has called a lot of attention in the field of the Industrial Internet of Things. Traditionally, Digital Twins were built as a software model of a physical entity, usually with the aid of CAD and modelling software technologies. But with the development of the Internet of Things and machine learning approaches, the Digital Twin concept will evolve and transform industrial operations

With the development of the Internet of Thing technologies, sensors connected to the physical asset may provide us with real-time data of the asset condition and send it back to complement the original digital twin with constant information of its actual performance. This continuous integration between the real-time data and the structural Digital Twin will allow us to manage systems more precisely than we now can imagine, optimizing the operation of industrial machines, from jet engines or manufacturing presses, to windmills.

Analytics plays a crucial role in building this updated version of the Digital Twin. On the one hand, capturing and plotting raw data with multiple dashboards is not feasible. Complex environments with multiple variables and parameters may challenge human ability to visually understand system behavior. On the other hand, setting thresholds and monitoring metrics against them may be useful in detecting simple anomalous situations, but the approach does not provide a digital representation of the asset behavior. More advanced analytics approaches are required to actually build a digital representation of the asset.

Novelti provides an autonomous behavior-monitoring system that allows us to build and update real-time views of the Digital Twin of an industrial asset. Novelti is an adaptive behavior-learning system that constructs and maintains a behavior repository – the behavioral representation of the Digital Twin. With this behavior-learning system, our users can monitor the real-time condition of the asset against the repository, and thus use this monitoring as an input for industrial operations and maintenance procedures.

The Novelti analytics capabilities allow users to adapt the asset representation and the operational procedures as the behaviors evolve, with early detection of anomalous situations and with very limited human intervention, providing the required efficiency to scale the digital representation to larger and more complex asset networks. In this article, we provide a more detailed description on how Novelti technology works and what makes it particularly suited of Industrial IoT environments.