Návrh systému pro kontrolu a vizualizaci dat z průmyslového robota

Abstract

This master's thesis deals with the design and practical implementation of a modular IIoT system for the collection, archiving, and visualization of operational and process data from an industrial robot. The research section analyzes current approaches to data acquisition from industrial robots, an overview of communication protocols, and available open-source and commercial platforms. The analytical section compares two technical implementation concepts and, based on this comparison, selects a modular stack consisting of Node-RED, InfluxDB, and Grafana. The practical section describes the implementation of the communication interface with the Universal Robots UR3e control unit via the Real-Time Client Interface, the configuration of the database for high-frequency time series archiving, and the design of a visualization dashboard. The system includes a feedback control loop implemented via the Modbus TCP protocol, enabling a response to detected process anomalies. The functionality of the system is verified through an experimental robotic screwing task with thread stripping detection, in which the torque profiles of standard and faulty assembly cycles are compared.