Claude Baron

Elie Saad, Aya Mrabah, Mariem Besbes et al.

Component obsolescence poses significant challenges in industries reliant on electronic components, causing increased costs and disruptions in the security and availability of systems. Accurate obsolescence risk prediction is essential but hindered by a lack of reliable data. This paper proposes a novel approach to forecasting obsolescence risk using zero-shot learning (ZSL) with large language models (LLMs) to address data limitations by leveraging domain-specific knowledge from tabular datasets. Applied to two real-world datasets, the method demonstrates effective risk prediction. A comparative evaluation of four LLMs underscores the importance of selecting the right model for specific forecasting tasks.

Elie Saad, Mariem Besbes, Marc Zolghadri et al.

The challenge of electronic component obsolescence is particularly critical in systems with long life cycles. Various obsolescence management methods are employed to mitigate its impact, with obsolescence forecasting being a highly sought-after and prominent approach. As a result, numerous machine learning-based forecasting methods have been proposed. However, machine learning models require a substantial amount of relevant data to achieve high precision, which is lacking in the current obsolescence landscape in some situations. This work introduces a novel framework for obsolescence forecasting based on deep learning. The proposed framework solves the lack of available data through deep generative modeling, where new obsolescence cases are generated and used to augment the training dataset. The augmented dataset is then used to train a classical machine learning-based obsolescence forecasting model. To train classical forecasting models using augmented datasets, existing classical supervised-learning classifiers are adapted for semi-supervised learning within this framework. The proposed framework demonstrates state-of-the-art results on benchmarking datasets.

Yann Argotti, Claude Baron, Phillipe Esteban

Nowadays, quality definition, assessment, control and prediction cannot easily be missed in systems engineering. One common factor among these activities is quality quantification. Therefore, throughout this paper, the authors focus on the problems relating to quality quantification in systems engineering. They first identify the main drawbacks of the current approaches adopted in this domain. They demonstrate how current solutions are not easily repeatable and adaptable across systems and how in most cases, the related standards such as ISO/IEC 25010 or Automotive-SPICE to cite just a few, are not used as they are within companies today. Fortunately, qualimetry, a young science with the purpose of quality quantification, provides the tools to resolve these gaps. To be able to use these tools, the authors propose a synthetic representation of qualimetry and its six pillars, named the ''House of Qualimetry'' and explain the fundamendal aspects of qualimetry. They identify a set of 8 attributes to characterize the design quality model and based on these attributes, propose a new process to design or adapt the quality model. Among these attributes, a new one is introduced to capture and measure the quality model evolution and adaptation aspect: the polymorphism and the polymorphism degree. Finally, the authors consolidate the measurement part thanks to a new measurement process before returning to the benefits of these contributions to systems engineering.

Diego Armando Diaz Vargas, Rui Xue, Claude Baron et al.

Agile methods are receiving a growing interest from industry and these approaches are nowadays well accepted and deployed in software engineering. However, some issues remain to introduce agility in systems engineering. The objective of this paper is to show an agile management implementation in an educational project consisting in developing a connected mobile robot, and to evaluate the issues and benefits of adopting an agile approach. Among the most famous agile management methods, SCRUM has been chosen to lead this experiment. This paper first presents the project and how students traditionally manage it, then it describes how Scrum could be used instead. It evaluates the difficulties and interests to introduce agility in this project, and concludes on the ability of Scrum to design, test and progressively integrate the system, thus providing an operational prototype more quickly.