The Industry 4.0 Concept assumes that the majority of industry’s resources will be able

to self-diagnose; this will, therefore, enable predictive maintenance. Numerically controlled

machines and devices involved in technological processes should, especially, have the facility

to predict breakdown. In the paper, the concept of a predictive maintenance system for

a vacuum furnace is presented. The predictive maintenance system is based on analysis of the

operating parameters of the system and on the algorithms for identifying emergency states in

the furnace. The algorithms will be implemented in the monitoring sub-system of the furnace.

Analysis of the operating parameters of vacuum furnaces, recorded in the Cloud will lead to

increased reliability and reduced service costs. In the paper, the research methodology for

identification of the critical parameters of the predictive maintenance system is proposed.

Illustrated examples of the thermographic investigation of a vacuum furnace are given.