Automated Configuration of Robots and Analytics in I4.0 with Digital Twins
The demonstrator for the automated configuration of robots and analytics processes shows how semantic digital twins can be employed in a plug-and-produce system to enable flexible small-batch manufacturing that conforms to Industry 4.0. The robot programs are automatically derived from the definitions of the production goals.
The digital twin model comprises semantic descriptions of processes, products and manufacturing resources. The underlying formalism of the semantic representation is a description logic that permits the system to automatically validate the logical consistency of models and also derive implicit facts from explicitly modeled facts.
The semantic digital twin model of a device encompasses a formal skill model based for example on information models and so-called companion specifications coded in OPC UA node sets. Standardized definitions for robot skills and their tools enable the manufacturer-independent specification of process steps.
Semantic skill models ensure semantic interoperability between the hardware and software components and between the production steps. They define the functional scope and applicability of a specific skill and as a result can be compared against the requirements of the individual production steps. Raw sensor data generated during the execution of the processes is automatically enriched with semantic context information in order to improve analytics processes based on machine learning.