Data Processing and Automation from Tightening Control Units Across Škoda Auto Plants

Abstract

This research establishes a pioneering cross-layer observability framework that transforms distributed industrial data into high-fidelity operational intelligence. While traditional automotive assembly relies on localized monitoring. This thesis designs a central analytical platform that connects different types of data streams from different production plants. The study uses Splunk's Search Processing Language (SPL) to show a complex system-level architecture that brings together infrastructure-level logs, production events, and communication reliability metrics into a single "source of truth." The primary innovation lies in the development of a rule-based anomaly detection capable of identifying systemic irregularities such as communication bottlenecks and tool-level degradation in near real-time. By using tiered, role-specific dashboards, the framework gives the production, maintenance, and IT teams synchronized information that makes diagnosis much easier and the process more clear. The results show that digital transformation across an entire organization can be done without changing the way control logic works. This was confirmed by using real-world production data. This work not only improves the efficiency of current operations, but it also lays the groundwork for predictive maintenance and advanced machine learning in discrete manufacturing. This puts it at the cutting edge of modern industrial research.