Product dimensional variability is a crucial factor in the quality control of complex multistage manufacturing processes, where undetected defects can easily be propagated downstream. The recent advances in information technologies and consequently the increased volume of data that has become readily available provide an excellent opportunity for the development of automated defect detection approaches that are capable of extracting the implicit complex relationships in these multivariate data-rich environments. In this paper, several machine learning classifiers were trained and evaluated on varied metrics to predict dimensional defects in a real automotive multistage assembly line. The line encompasses two automated inspection stages with several human-operated assembly and pre-alignment stages in between. The results show that non-linear models like XGBoost and Random Forests are capable of modelling the complexity of such an environment, achieving a high true positive rate and showing promise for the improvement of existing quality control approaches, enabling defects and deviations to be addressed earlier and thus assist in reducing scrap and repair costs.

No abstract

No abstract

Industrial Cyber-Physical Systems (CPS) are promoting the development of smart machines and products, leading to the next generation of intelligent production systems. In this context, Artificial Intelligence (AI) is posed as a key enabler for the realization of CPS requirements, supporting the data analysis and the system dynamic adaptation. However, the centralized Cloud-based AI approaches are not suitable to handle many industrial scenarios, constrained by responsiveness and data sensitive. Edge Computing can address the new challenges, enabling the decentralization of data analysis along the cyber-physical components. In this context, distributed AI approaches, such those based on Multi-agent Systems (MAS), are essential to handle the distribution and interaction of the components. Based on that, this work uses a MAS approach to design cyber-physical agents that can embed different data analysis capabilities, supporting the decentralization of intelligence. These concepts were applied to an industrial automobile multi-stage production system, where different kinds of data analysis were performed in autonomous and cooperative agents disposed along Edge, Fog and Cloud computing layers.Industrial Cyber-Physical Systems (CPS) are enabling the next generation of intelligent production systems, mainly based on the concepts of smart machines and products. Driven by the needs to attend the ever-changing market trends, such digital transformation is mainly based on the use of Internet of Things (IoT), Cloud Computing and Artificial Intelligence (AI) technologies [12]. While the first enables the interconnection of equipment and consequently the digitization of the industrial environment [22], the second provides on demand high processing and storage resources [15]. On the other hand, AI provides advanced data analysis algorithms, such those based on Machine-Learning (ML), that can take advantage of the huge amounts of IoT data and the power of Cloud Computing, in order to provide actionable information and support data-driven decision-making [20,8].Although Cloud manufacturing [15] has been seen as a new paradigm in the realization of the 4th industrial revolution (4IR) [12], the traditional Cloud-based approaches, where IoT devices send all the data to be processed by centralized applications, present some drawbacks. Indeed, besides information security and privacy concerns [21], this approach is not suitable for many real-time, data-sensitive and constrained network applications [2]. In this context, Fog Computing emerged to cover the Cloud limitations, promoting the deployment of data processing capabilities closer to the data sources [4]. It defines an intermediate computing layer between Cloud applications and IoT devices that besides providing a more direct, reliable, secure and fast link between them, also promotes the decentralization of data analysis, decision-making and control, increasing local components autonomy.Besides Fog, which considers equipment at the local network, CPS also considers processing ca...

No abstract

One of the areas that can heavily benefit with Industry 4.0 is the logistics, namely with the association of sensing technologies and the application of techniques such as Big Data Analytic, Data Visualization, prediction algorithms, and especially 3D simulation. The association of real data, prediction techniques, and 3D models, allow the creation of realistic Digital Twins that emulate factory processes, making possible the experimentation and testing of new ideas and different scenarios by tweaking key variables, without stopping production. However, there are many challenges in order to handle and compute all fast-growing, multi dimension data generated, so that all this production related data can be quickly used for defect control, preventive maintenance, advanced analytics for production and resources management, or even later simulation. The work presented in this paper focus in this “in between” processing work, presenting an easily deployable and self-reconfigurable Big Data architecture, where different technologies can work together to extract, transform, load, apply analytics, and then feed a 3D Digital Simulation model. The work presented in this paper is funded by the EU project BOOST4.0 and focus in a specific logistic process of car manufacturing.

This paper presents the design and the characterization of a portable laser triangulation measurement system for measuring gap and flush in the car body assembly process. Targeting Human in the Loop (HILT) operations in the manufacturing sector, and in line with the vision of human empowerment with Industry 4.0 technologies, the instrument embeds features to ease operators’ activity and compensate possible misuse that could affect the robustness and the quality of data acquired. The device is based on a smartphone integrated with a miniaturized laser triangulation system installed in a cover. The device embodies additional sensors and control systems in order to guarantee operators’ safety (switching on and off the laser line based on specific conditions), support operators during the measurement execution task, and optimize the image acquisition process for minimizing the uncertainty associated to the measurement. The smartphone performs on-board processing and allows Wi-Fi communication with the plant IT infrastructure. Compliance to Industry 4.0 requirements is guaranteed using OPC-UA (Open Platform Communications—Unified Architecture) communication protocol enabling the exchange of live data with the plant middleware. The smartphone provides also an advanced high-resolution color display and well proven and ergonomic human–machine interfaces, which have been fully exploited in the design. The paper introduces the system optical layout and then presents the algorithms implemented to realize the gap and flush measurement. The paper finally presents the calibration of the instrument and estimates its calibration uncertainty in laboratory conditions. Then it discusses how performance decays when the operator handles the instrument on a reference car body. Finally, it shows the analysis of uncertainty when the device is used on real car bodies of different colors in a production line. It is observed that the measurement uncertainty of the whole measurement chain (measurand + instrument + operator + uncontrolled environmental conditions) is larger than the instrument calibration uncertainty because the measurement process is affected by the operator and the variable conditions of the production line.

No abstract