Abstract
Reliable operation of production lines without unscheduled disruptions is of paramount importance for ensuring the proper operation of automated working cells involving robotic systems. This article addresses the issue of preventing disruptions to an automotive production line that can arise from incorrect placement of aluminum car parts by a human operator in a feeding container with 4 indexing pins for each part. The detection of the misplaced parts is critical for avoiding collisions between the containers and a high pressure washing machine and also to avoid collisions between the parts and a robotic arm that is feeding parts to a air leakage inspection machine. The proposed inspection system relies on a 3D sensor for scanning the parts inside a container and then estimates the 6 DoF pose of the container followed by an analysis of the overlap percentage between each part reference point cloud and the 3D sensor data. When the overlap percentage is below a given threshold, the part is considered as misplaced and the operator is alerted to fix the part placement in the container. The deployment of the inspection system on an automotive production line for 22 weeks has shown promising results by avoiding 18 hours of disruptions, since it detected 407 containers having misplaced parts in 4524 inspections, from which 12 were false negatives, while no false positives were reported, which allowed the elimination of disruptions to the production line at the cost of manual reinspection of 0.27% of false negative containers by the operator.

Abstract
The NEXUS project involves innovation and entrepreneurship. The first task of the project team, which is eclectic and includes several different profiles (research assistants, researchers, managers, engineers, designers and marketing specialists) was to do a stakeholder analysis, which involved doing some field work – six unstructured interviews – as well as some secondary data analysis (scientific articles on the Scopus database; Google searches; among others). The main concerns were to: unlock value through stakeholder engagement; move from concept to collaboration – for sustainable port development; and to connect ports worldwide which will involve international stakeholder collaboration and synergies for sustainable innovation. A Stakeholder analysis is a process consisting of identifying the individuals or groups that have a specific interest in a project, which can affect them, or be affected by them, including large groups such as customers and suppliers; employees; shareholders and investors. It is necessary to sort them according to their impact and consequences regarding the project. It is essential to manage stakeholder expectations when making decisions, especially in a project with thirty-five partners and with a global reach. The initial funding is around 92 million euros (approved by the Recovery and Resilience Plan (RRP)), to ensure the ambitious implementation of the new NEXUS platform and other related products. To minimize backlash, early as well as ongoing stakeholder involvement is a prerequisite. Strategy formulation should not occur in an “ivory tower” and involvement is key. For effective and efficient decision-making, to minimize risks, which need to be identified as well as all the connections and relationships to be created and nurtured. The research question at the outset was: How may we successfully implement a multi-partner project which involves considerable investment, and which aims to digitalize maritime ports worldwide? The theme is, currently, without precedent as the impact intended is for there to be a positive impact on all stakeholders involved – a win-win situation all around. The project will finish in 31-12-2025 and will produce 28 innovative solutions for commercialization. NEXUS is a green and digital project which is up to date as regards the latest sustainability norms and regarding overall operational efficiency in multimodal logistics. Work Package 10 is a specific exploitation, internationalization, promotion, and dissemination package.

Abstract
Ensuring privacy while sharing sensitive data is critical, particularly in fields such as healthcare, and everywhere compliance with data protection regulations is required. Anonymization and pseudonymization techniques are essential for preserving individual privacy but it is challenging to select the most appropriate methods given particular privacy and utility requirements. We conducted a focus group during the EuroPLoP 2024 conference that aimed to obtain feedback on patterns that we documented in this space and on a pattern map we outlined, and to identify patterns related to anonymization or pseudonymization of data that have not yet been documented. Some of the patterns we documented were not known by participants. On the other hand, we found some techniques that are potentially privacy-preserving patterns that have not yet been documented, and framed these techniques according to the category in our pattern map. Although the results suggest that our current patterns address some recurring privacy challenges, further exploration and documentation of the techniques are necessary to capture the full range of privacy-preserving solutions.

Abstract

Abstract
This study investigates the effectiveness of Transformer-based models for retail demand forecasting. We evaluated vanilla Transformer, Informer, Autoformer, PatchTST, and temporal fusion Transformer (TFT) against traditional baselines like AutoARIMA and AutoETS. Model performance was assessed using mean absolute scaled error (MASE) and weighted quantile loss (WQL). The M5 competition dataset, comprising 30,490 time series from 10 stores, served as the evaluation benchmark. The results demonstrate that Transformer-based models significantly outperform traditional baselines, with Transformer, Informer, and TFT leading the performance metrics. These models achieved MASE improvements of 26% to 29% and WQL reductions of up to 34% compared to the seasonal Na & iuml;ve method, particularly excelling in short-term forecasts. While Autoformer and PatchTST also surpassed traditional methods, their performance was slightly lower, indicating the potential for further tuning. Additionally, this study highlights a trade-off between model complexity and computational efficiency, with Transformer models, though computationally intensive, offering superior forecasting accuracy compared to the significantly slower traditional models like AutoARIMA. These findings underscore the potential of Transformer-based approaches for enhancing retail demand forecasting, provided the computational demands are managed effectively.

Abstract
While a wide range of resources is available on orchestration techniques and best practices for containerized software systems, many are not documented clearly or in detail. This complicates the process of selecting the most suitable methods for various usage scenarios. To address this gap, we documented a set of orchestration patterns. This paper reports the results of a focus group conducted during the EuroPLoP 2024 conference, where we aimed to obtain feedback on that group of patterns and on a wider pattern map we outlined. We also aimed to identify container orchestration patterns that have not yet been documented. We found that participants knew most of the patterns we included on the pattern map. Additionally, one of the practices mentioned by the participants (Node Balancing) was previously documented as a pattern by us with the name of Service Balancing. Finally, we found important insights into container orchestration patterns, expanding our pattern map to include eight new proto-patterns.