Additive manufacturing represents one of the key technological pillars of industrial transformation in the context of Industry 4.0, where production quality, precision and repeatability depend not only on material selection and design, but primarily on the optimisation of process parameters and the implementation of advanced monitoring systems. FDM/FFF technologies, among the most widely used additive manufacturing methods, are highly sensitive to variations in printing conditions, which often result in different types of process defects. Continuous monitoring and early detection of such errors are therefore essential for improving production reliability and stability. This paper focuses on the analysis of typical process defects in FDM/FFF technology, with emphasis on their identification and prediction using intelligent monitoring tools based on artificial intelligence and IoT elements. An experimental monitoring system was developed using a Raspberry Pi 3 microcomputer, a Raspberry camera and the OctoPrint software platform extended with AI-based modules, particularly the Detector 2 plugin. The system was experimentally validated through a series of printing tests performed on a Creality CR-30 PrintMill printer, targeting defects such as stringing, layer deformation and material flow instabilities.