Industry 4.0 technologies in quality and safety control systems in food manufacturing: A systematic techno-managerial analysis on benefits and barriers

ABSTRACT Background The food industry faces increasing demands for improved quality and safety, while conventional quality control methods remain labour-intensive, slow, and limited. Industry 4.0 (I4.0) technologies show promise, but real-world implementation remains limited. Advancing practice requires clear insight into technical/technological and managerial benefits and barriers. Scope This review examines the I4.0 technologies and their implementation status in quality and safety systems in food manufacturing, and their applicability in either product or process quality control, as well as in elements of the quality control circle (data collection and analysis, corrective and proactive actions). Followingly, the benefits and barriers of these technologies that are mentioned in the reviewed studies are categorised using a techno-managerial approach. Key Findings and Conclusions Artificial intelligence (AI) is mainly used for product quality control, while the Internet of Things supports process quality control in the reviewed studies. Data analysis is the most addressed element of the quality circle; AI has the most potential. The reported benefits are primarily technical/technological, focusing on contamination detection and real-time quality monitoring. Managerial benefits, though less emphasised, include cost-effectiveness, better food safety and crisis management, and strategic improvement. Key technical/technological barriers are process and equipment-related , notably the need for high-quality data and time-intensive AI model training for large or complex datasets. Besides, reliable and accurate performance can still be a barrier due to overfitting, misclassification, etc. Managerial barriers are mostly people-related , including manual labelling errors and security issues. A multidisciplinary approach is essential to overcoming these barriers and promoting field implementations.