Advanced manufacturing is undergoing a profound transformation, with data quickly becoming its most strategic asset. The industry is pushing toward Industry 4.0 with its sights already on the human-centric Industry 5.0. Manufacturing firms are rapidly integrating AI, IoT, and advanced analytics to enable real-time decision making, predictive maintenance, and full manufacturing lifecycle optimization. However, this data-driven revolution exposes a critical vulnerability: the hidden direct costs and cascading downstream consequences of inaccurate, missing, or corrupt data. This paper provides an in-depth examination of the data quality crisis facing modern manufacturing, exploring its quantifiable impact on cost, safety, and strategic decision making; and identifies the tangible barriers preventing scalable AI in manufacturing today. We investigate how bad data undermines the digital thread, erodes both operational and strategic trust, and stalls the transition to autonomous systems. Supported by recent industry surveys, academic findings, and leading trends, we reveal that most manufacturers suffer from systemic data quality issues, with billions lost annually to inefficiencies, rework, and flawed decisions. Addressing this, the paper evaluates state-of-the-art solutions for real-time data validation, anomaly detection, and predictive imputation. Building upon this, we identify key gaps—including the lack of unified data quality frameworks, integration across legacy/modern systems, and actionable imputation under uncertainty—and propose a roadmap to bridge them. The paper concludes by outlining four research directions that support a seamless, scalable transition toward a trustworthy data foundation in manufacturing. Industry 4.0/5.0 is defined by data, insight, and actionable intelligence: only manufacturers that tame their data chaos will thrive.

This paper presents a hybrid model that combines Artificial Neural Networks (ANN) and Gaussian Processes (GP). The goal is to achieve high prediction accuracy while quantifying uncertainty. The proposed structure is a simple ANN used as the trend of the GP, particularly emphasizing the joint training of the two models. The ANN+GP exploits the ability of the ANN to capture complex, non-linear relationships in the data. At the same time, the GP provides an approach to uncertainty estimation, thus improving the accuracy of the predictions. This paper emphasizes the importance of concurrent training, which can improve the accuracy of the prediction model. The algorithm is suitable for any application where both accurate, robust predictions and uncertainty estimates are critical to enhance the interpretability of the model. The proposed method has been successfully applied to the frequency response functions of a simple structure.