Návrh systému včasného varování pro identifikaci krizových situací v podniku

Abstract

This thesis focuses on the design of an early warning system for identifying crisis situations in a medium-sized manufacturing company operating in the plastics industry. The main objective of the thesis is to analyze the weaknesses of the existing reactive risk detection system and propose a proactive solution that will reduce the information delay in top management decision-making. In the practical part, a mixed-methods approach was used, including semi-structured interviews and quantitative analysis of company data. The results of the investigation revealed a barrier in the form of isolated information silos. Subsequent regression analysis statistically confirmed a moderately strong causal relationship between production capacity overload and an increase in employee absenteeism, with r = 0.533. Based on these findings, an automated EWS dashboard was designed in the Microsoft Power BI environment. This system innovatively integrates operational data from machines via the OPC UA protocol with the employee attendance system. A subsequent FMEA analysis demonstrated that implementing this solution radically increases threat detectability and reduces the level of monitored risks by more than 60 %. The economic evaluation ultimately confirmed the high profitability of the proposed project, with an ROI of 188.5 % and a payback period of 4.2 months.