An industrial plant machine learning system includes a machine learning model, providing machine learning data, an industrial plant providing plant data and an abstraction layer, connecting the machine learning model and the industrial plant, wherein the abstraction layer is configured to provide standardized communication between the machine learning model and the industrial plant, using a machine learning markup language.

Various embodiments of the present technology relate to digital twins of devices and assemblies. More specifically, some embodiments relate to the orchestration of digital twin models for representing industrial systems based on characteristics of digital twins. In an embodiment, a method of operating an orchestration engine in an industrial automation environment comprises identifying a targeted outcome for modeling the industrial automation environment, configuring a digital twin environment corresponding to the industrial automation environment based at least on the targeted outcome, and executing a process associated with the industrial automation environment using the digital twin environment.

In a method for optimizing a modular system for technical functional units of a process engineering plant, a modular system is provided having components for configuring technical functional units. The modular system can be projected in a simulation environment such that each component can be represented based on its physical properties as a virtual component with corresponding parameters in the simulation environment. Parameters of the virtual components are varied in the simulation environment to determine a modified configuration of at least one technical functional unit using at least one of the virtual components with at least one varied parameter. The operation of the technical functional unit(s) is simulated with the modified configuration. A set of virtual components is determined from the virtual components with varied parameters based on results of the simulation. One or more components of the modular system are adapted based on the determined set of virtual components.

A modular modeling engine is provided for industrial automation applications. The module may be instantiated upon demand, such as upon receipt of annotated data for a system or process being monitored and/or controlled. The model is agnostic insomuch as little or no prior knowledge is required of the system or process. Variables, functions, and their combinations are selected and the model is refined automatically. A data structure is received for instantiation of the model, and following modeling, a similar data structure is produced. The module may be used together with other modules for caning out complex automation processing at the same or multiple levels in an automation setting.

A process model is created from site data generated in a process.

A process model creation system includes: a site data acquiring section that acquires, as site data, task information related to one or more tasks implemented for a process including the tasks, and task-related information related to the implemented tasks; an instance creating section that creates a step instance by associating the task-related information with the corresponding task information; and a process model creating section that creates, from one or more step instances, a process model which is a model representing a relationship between the one or more tasks in the process.

A method of hierarchical machine learning includes receiving a topology model having information on hierarchical relations between components of the industrial plant, determining a representation hierarchy comprising a plurality of levels, wherein each representation on a higher level represents a group of representations on a lower level, wherein the representations comprise a machine learning model, and training an output machine learning model using the determined hierarchical representations.

A method of transfer learning for a specific production process of an industrial plant includes providing data templates defining expected data for a production process, and providing plant data, wherein the data templates define groupings for the expected data according to their relation in the industrial plant; determining a process instance and defining a mapping with the plant data; determining historic process data; determining training data using the determined process instance and the determined historic process data, wherein the training data comprises a structured data matrix, wherein columns of the data matrix represent the sensor data that are grouped in accordance with the data template and wherein rows of the data matrix represent timestamps of obtaining the sensor data; providing a pre-trained machine learning model using the determined process instance; and training a new machine learning model using the provided pre-trained model and the determined training data.

A system and method for generating a holistic digital twin of an industrial facility that includes a plurality of assets, wherein the said system includes conversion units configured to convert asset related data collected from different tools utilized to plan and/or operate the industrial facility in a tool specific data format and asset related data provided by data sources of the industrial facility in a data source specific data format into a common graphical representation, a matching unit configured to match the common graphical representations of the converted asset related data to provide a mapping between the assets of the industrial facility, and a merging unit configured to merge the mapped assets of the industrial facility into a unified graph to provide the holistic digital twin of the industrial facility.

An industrial CAD system is supplemented with features that allow a developer to easily convert a mechanical CAD model of an automation system to a dynamic digital twin capable of simulation within a simulation platform. The features allow the user to label selected elements of a mechanical CAD drawing with “aspects” within the CAD environment. These aspect markups label the selected mechanical elements as being specific types of industrial assets or control elements. Based on these markups, the CAD platform associates mechatronic metadata with the selected elements based on the type of aspect with which each element is labeled. This mechatronic metadata defines the behavior (e.g., movements, speeds, forces, etc.) of the selected element within the context of a virtual simulation, transforming the mechanical CAD model into a dynamic digital twin that can be exported to a simulation and testing platform.

Generating a digital representation of a process automation system on a cloud-based service platform uses assets integrated into measurement points. The method includes reading TAG information using an edge device, wherein the TAG information is provided in a character chain data type and represents the hierarchical structure of the respective asset. The method also includes transmitting the TAG information to the cloud-based service platform, and parsing the TAG information using an application, wherein a logic is used for the parsing process, and the name of the asset and the name of the measurement point in which the respective asset is integrated are extracted from the TAG information. A structure plan of the system is generated using the application having all of the system measurement points extracted from the TAG information together with all of the assets which are assigned to the measurement points and are extracted from the TAG information.