Resistance spot welding is a process with high variability regarding the quality of the produced joints. This means that key performance indicators (KPIs) such as geometrical and mechanical features as well as failure modes can deviate from the initial design even if the same process parameters are used. The industry has developed quality assurance programs and quality control methods for tracking these KPIs; however, most of them are based on offline or/and destructive practices. On the other hand, state-of-the-art approaches have made online quality assessment feasible and proved its necessity if a 100% quality rate is required. However, limited attention has been given to “closing the loop” and providing feedback for preventing and correcting process anomalies that cause quality variations in real time. In this study, the main gaps between the state of practice and the state of the art are discussed in the context of quality assurance for resistance spot welding. Finally, the role and importance of digital twins by taking into consideration the entire welding ecosystem in quality assurance are discussed in order to form the prospects for the road ahead.

The automation of workflows for the optimization of manufacturing processes through digital twins seems to be achievable nowadays. The enabling technologies of Industry 4.0 have matured, while the plethora of available sensors and data processing methods can be used to address functionalities related to manufacturing processes, such as process monitoring and control, quality assessment and process modelling. However, technologies succeeding Computer-Integrated Manufacturing and several promising techniques, such as metamodelling languages, have not been exploited enough. To this end, a framework is presented, utilizing an automation workflow knowledge database, a classification of technologies and a metamodelling language. This approach will be highly useful for creating digital twins for both the design and operation of manufacturing processes, while keeping humans in the loop. Two process control paradigms are used to illustrate the applicability of such an approach, under the framework of certifiable human-in-the-loop process optimization.

Electric vehicles’ batteries, referred to as Battery Packs (BPs), are composed of interconnected battery cells and modules. The utilisation of different materials, configurations, and welding processes forms a plethora of different applications. This level of diversity along with the low maturity of welding designs and the lack of standardisation result in great variations in the mechanical and electrical quality of the joints. Moreover, the high-volume production requirements, meaning the high number of joints per module/BP, increase the absolute number of defects. The first part of this study focuses on associating the challenges of welding application in battery assembly with the key performance indicators of the joints. The second part reviews the existing methods for quality assurance which concerns the joining of battery cells and busbars. Additionally, the second part of this paper identifies the general trends and the research gaps for the most widely adopted welding methods in this domain, while it renders the future directions.

Vision technologies are used in both industrial and smart city applications in order to provide advanced value products due to embedded self-monitoring and assessment services. In addition, for the full utilization of the obtained data, deep learning is now suggested for use. To this end, the current work presents the implementation of image recognition techniques alongside the original the quality assessment of a Parabolic Trough Collector (PTC) reflector surface to locate and identify surface irregularities by classifying images as either acceptable or non-acceptable. The method consists of a three-step solution that promotes an affordable implementation in a relatively small time period. More specifically, a 3D Computer Aided Design (CAD) of the PTC was used for the pre-training of neural networks, while an aluminum reflector surface was used to verify algorithm performance. The results are promising, as this method proved applicable in cases where the actual part was manufactured in small batches or under the concept of customized manufacturing. Consequently, the algorithm is capable of being trained with a limited number of data.

For smart manufacturing systems, quality monitoring of welding has already started to shift from empirical modeling to knowledge integration directly from the captured data by utilizing one of the most promising Industry 4.0’s key enabling technologies, artificial intelligence (AI)/machine learning (ML). However, beyond the advantages that they bring, AI/ML introduces new types of security threats, which are related to their very nature and eventually, they will pose real threats to the production cost and quality of products. These types of security threats, such as adversarial attacks, are causing the targeted AI system to produce incorrect or malicious outputs. This may undermine the performance (and the efficiency) of the quality monitoring systems. Herein, a software platform servicing quality monitoring for welding is presented in the context of resistance and submerged arc welding. The hosted ML classification models that are trained to perform quality monitoring are subjected to two different types of untargeted, black-box, adversarial attacks. The first one is based on a purely statistical approach and the second one is based on process knowledge for crafting these adversarial inputs that can compromise the models’ accuracy. Finally, the mechanisms upon which these adversarial attacks are inflicting damage are discussed to identify which process features or defects are replicated. This way, a roadmap is sketched toward testing the vulnerability and robustness of an AI-based quality monitoring system.