ADAPT: an Automatic Diagnosis of Activity and Processes in auTomation environments

Abstract

In an automated industrial environment, a large volume of data and signals is available, both from sensors and actuators in machinery and from the interaction with operators and users. Operation diagnosis can have multiple applications from a learning point of view (e.g. staff training) or in terms of process assessment. This work proposes a methodology for implementing an Intelligent System by means of any interactive system connected through OPC UA standard. A digital twin of the process supports configuration and validation, prior to commissioning. Activity is interpreted and diagnosed according to the context in which it occurs. Step order in sequence, step duration and sequence duration are analyzed in a use case based on a PLC-controlled robotic cell in which it is operated both in automatic and manual mode for adjusting a linear table's positioning.