Object detection and other machine learning technology applications play an important role in various areas of the product lifecycle, especially in quality assurance or general assembly assistance. While the implemented computer vision-based systems provide great benefits, training and implementing deep learning models is often a tedious and time-consuming task, especially in the field of ob-ject detection. To accomplish good results, large datasets with a high quantity of object instances in a bright variety of poses are required These are generally created manually and are therefore very time consuming to create.

To improve the training process, synthetic training data can be used. It is generated within a virtual environment using a product’s geometry model. In this paper, the authors propose a synthetic dataset generator for object detection, that is integrated into a PLM system to automate the process of collecting and processing the CAD data for creating the synthetic machine learning training dataset. Domain randomization is used to eliminate the effort of creating a virtual environment, to fully automate the dataset generation, and to to increase the generalization of the model. The trained detector is tested on an object detection demonstrator set-up to evaluate its performance in a real-world use case. For evaluation purposes, the authors also provide a comparison of the test results to an object detection model that is trained without domain randomization, using a very close-to-reality virtual environment.

Digital technologies enables industries to transform their processes to gain competitive advantage. Industry 5.0 puts the operator at the center of a digital and connected industry, but what about workers' emotions? To what extent do Industry 4.0 technologies in the industrial domain allows for a better understanding of the impact of emotions on workers' decision-making behavior? The overall objective of this systematic literature review is to explore the literature to assess the breadth of possibilities for analyzing emotions to understand workers' decision-making behaviors, based on data collected in industrial settings.

The analysis of 29 articles extracted from the Compendex and Web of Science search engines allowed us to define the emotional factors measured in the analysis of human decision-making behavior, the tools used, and the sectors of application. The subject is still in its infancy for the scientific community and is a source of excitement.

The results of the qualitative analysis of the articles show the predominance of text analysis (social networks and/or online reviews) for sentiment analysis. The tools used within the technology are very diverse (deep learning, machine learning, mathematical models). The same is true for the sectors of activity, although there is a particular interest in customer emotions for marketing purposes in the service industries. Finally, future research avenues are proposed, such as the analysis of the impact of emotions on the decision-making process in manufacturing is practically absent from the study.

Business Process Mining is considered one of the merging fields that focusses on analyzing Business Process Models (BPM), by extracting knowledge from event logs gener-ated by various information systems, for the sake of auditing, monitoring and analysis of business activities for future improvement and optimization throughout the entire lifecycle of such processes, from creation to conclusion. In this work, Long Short-Term Memory (LSTM) Neural Network was utilized for the prediction of the execution of cases, through training and testing the model on event traces extracted from event logs related to a given business process model. From the initial results we obtained, our model was able to predict the next activity in the sequence with high accuracy. The approach consisted of three phas-es: preprocessing the logs, classification, and categorization and all the activities related to implementing the LSTM model, including network design, training, and model selection. The predictive analysis achieved in this work can be extended to include anomaly detection capabilities, to detect any anomalous events or activities captured in the event logs.

In smart manufacturing, data-driven artificial intelligence algorithms are be-coming increasingly important in improving decision-making by monitoring the control, analysis, and prediction of manufacturing processes in a produc-tion system. In the textile industry, there is a strong need for smart manufac-turing technologies because various parameters could affect the quality dy-namically. This study aims to optimize the parameters of spinning processes by developing machine learning algorithms and models which can predict the toughness and elasticity of threads. At first, meaningful variables are extract-ed from the shop floor data, and then a defect classification learning model is developed to predict defects in advance. In addition, a regression model is implemented for the prediction of toughness and elasticity of the textile. By transitioning from the traditional trial and error method to the data-based method for the spinning process, production costs and time can be reduced through optimal settings of the production parameters for the spinning of the desired threads.