ESI Group (France)

An interpretable LiDAR-based architecture is proposed for the detection and classification of weather conditions in the context of automated driving. The approach integrates Gray-Level Co-occurrence Matrix (GLCM) texture features with explainable artificial intelligence (XAI) techniques to enhance both performance and interpretability. Exploiting the sensitivity of LiDAR to atmospheric particles, the system is designed to identify the most common weather scenarios in Europe: clear, foggy, and rainy conditions. Raw LiDAR point clouds are pre-processed and projected onto a discretized bird’s-eye view (BEV) map to compute GLCMs, which serve as input to a lightweight feedforward neural network. The use of GLCM not only facilitates classification but also supports interpretability by providing structured texture representations of the scene. Two complementary XAI approaches are employed. First, an ante-hoc statistical analysis estimates the conditional probability distributions of GLCM features given each weather class. Second, a learned weight matrix provides a visual representation of the features captured by the model. The proposed model is trained and evaluated on the publicly available RADIATE dataset and benchmarked against three state-of-the-art weather classification methods. Experimental results demonstrate that the architecture achieves competitive accuracy while maintaining low computational complexity, making it suitable for real-time embedded applications in autonomous driving systems.

Introdução: A integração da sustentabilidade no ensino da fisioterapia constitui uma resposta urgente aos desafios da saúde planetária e da responsabilidade social dos profissionais de saúde. Contudo, permanece limitada a investigação sobre o nível de preparação dos estudantes nesta área. Este estudo teve como objetivo explorar atitudes, competências percebidas, comportamentos e barreiras relativamente à sustentabilidade entre estudantes de fisioterapia de uma instituição de ensino superior em Portugal. Metodologia: Foi aplicado um questionário online, intitulado Atitudes e competências em sustentabilidade em Fisioterapia a 113 estudantes (idade média = 22 anos; 60,2% do género feminino), de todos os anos curriculares, de uma escola de ensino superior na Licenciatura em Fisioterapia. O mesmo foi previamente aprovado pela comissão de ética da mesma instituição com o parecer 2025/05-06. Resultados: Apenas 13,3% referiram ter recebido formação prévia em sustentabilidade ou saúde ambiental. A maioria reconhece a ligação entre alterações ambientais e saúde (81,4%), considera importante a inclusão do tema nos planos curriculares (67,3%) e adota comportamentos sustentáveis, como a redução de materiais descartáveis (61,9%). No entanto, foram identificadas barreiras como a falta de formação (49,6%) e o apoio institucional limitado (44,2%). Foram encontradas diferenças estatisticamente significativas entre géneros nas secções “Atitudes e Crenças” (χ²(4)=13,4, p=0,010) e “Barreiras Percebidas” (χ²(4)=13,1, p=0,011), indicando perceções diferenciadas nestas dimensões. A comparação entre estudantes com e sem formação prévia não revelou diferenças significativas em termos de conhecimento percebido (t(111) = -0,577; p=0,565), sugerindo uma fragilidade na profundidade ou eficácia da formação existente. Discussão: Os resultados revelam uma atitude globalmente favorável à integração da sustentabilidade na fisioterapia, mas também uma lacuna formativa e institucional que dificulta a sua concretização prática. Conclusão: Este estudo reforça a necessidade de estratégias educativas mais consistentes e interdisciplinares, capazes de capacitar futuros fisioterapeutas para uma prática crítica, ambientalmente responsável e orientada para a saúde planetária.

Although growing operational automation and informational integration are having a positive impact on ports’ competitiveness, they come with increased cyber risks. Fed by data collected from 18 French ports’ websites and associated internet sources, a study has been conducted to assess these ports’ level of cyber risk awareness and identify their possible cybersecurity initiatives. Results show that the sampled ports are still on their way towards full cyber risk awareness and global cyber risk management, but they are making significant progress in the right direction.

<div class="section abstract"><div class="htmlview paragraph">Permanent magnet synchronous motors (PMSM) are among the most promising motors in electric vehicles due to their high torque density and efficiency. This paper is devoted to detailed electromagnetic investigations of permanent magnet synchronous motor, accounting for specific rotor eccentricity and uneven magnetization. A series of simulations are performed for a 90 HP interior PMSM to investigate the changes in the radial and tangential forces when the rotor is perfectly aligned or with static, dynamic, and mixed eccentricities. Besides, the influence of uneven magnetization due to manufacturing, demagnetization, and magnet deterioration is discussed. The forces are then used to load a vibro-acoustic model to evaluate the impact on the noise, vibration, and harshness (NVH) performance and predict the radiated sound power level for the different conditions.</div></div>

The original version of this Article contained an error in Fig. 5 , in which the colour coded precipitation range legend contained wrong values. This has been corrected in both the PDF and HTML versions of the Article.

In aeroacoustic engineering, replicating real-world environments in a controlled setting presents significant opportunities for reduced flight testing but also comes with challenges when precisely simulating the flight environment. This paper introduces an innovative acoustic simulation model that leverages the Boundary Element Method (BEM) and a Direct Field Acoustic Testing (DFAT) simulation module. The primary objective is to virtually reproduce test sound fields that closely mimic the full fluctuating pressure environments aircraft encounter during flight using only loudspeaker arrays. This approach involves virtually reproducing a sophisticated test setup, typically comprising hundreds of drivers actively managed to generate a specific sound field. This setup is designed to virtually replicate various complex pressure environments, such as those produced by a Turbulent Boundary Layer, the complex pressure fields generated by aircraft propellers or jet engines. The use of active noise generation simulation helps determine the feasibility of a given pressure field and the capability of a test facility to simulate it accurately. The virtually reproduced sound fields simulated by this model show remarkable potential in approximating real-world pressure loadings. This capability is particularly useful in establishing pre-test predictions for fuselage testing such as sidewall build-ups: including insulation, and damping treatments, TVAs and TMDs.

We develop a novel methodology for data-driven simulation of coupled multi-physics systems. The result of the method is a learned numerical integrator of the coupled system dynamics. In order to preserve the fundamental physics of the coupled systems, and thus preserve the geometrical properties of the governing equations—even if they may be completely unknown—we impose a port-metriplectic structure on the system evolution, i.e., a combination of a symplectic evolution for the system energy with a gradient flow for the entropy of each system, which can be exchanged through predefined ports. The resulting method guarantees by construction the satisfaction of the laws of thermodynamics for open systems, leading to accurate predictions of the future states of their dynamics. Examples are given for systems of varying complexity, based on synthetic as well as experimental data.

Relaxing the sequential specification of a shared object is a way to obtain an implementation with better performance compared to implementing the original specification. We apply this approach to the Counter object, under the assumption that the number of times the Counter is incremented in any execution is at most a known bound $m$. We consider the $k$-multiplicative-accurate Counter object, where each read operation returns an approximate value that is within a multiplicative factor $k$ of the accurate value. More specifically, a read is allowed to return an approximate value $x$ of the number $v$ of increments previously applied to the counter such that $v/k \le x \le vk$. We present three algorithms to implement this object in a wait-free linearizable manner in the shared memory model using read-write registers. All the algorithms have read operations whose worst-case step complexity improves exponentially on that for an exact $m$-bounded counter (which in turn improves exponentially on that for an exact unbounded counter). Two of the algorithms have read step complexity that is asymptotically optimal. The algorithms differ in their requirements on $k$, step complexity of the increment operation, and space complexity.

The rapid growth of cultural evolutionary science, its expansion into numerous fields, its use of diverse methods, and several conceptual problems have outpaced corollary developments in theory and philosophy of science. This has led to concern, exemplified in results from a recent survey conducted with members of the Cultural Evolution Society, that the field lacks 'knowledge synthesis', is poorly supported by 'theory', has an ambiguous relation to biological evolution and uses key terms (e.g. 'culture', 'social learning', 'cumulative culture') in ways that hamper operationalization in models, experiments and field studies. Although numerous review papers in the field represent and categorize its empirical findings, the field's theoretical challenges receive less critical attention even though challenges of a theoretical or conceptual nature underlie most of the problems identified by Cultural Evolution Society members. Guided by the heterogeneous 'grand challenges' emergent in this survey, this paper restates those challenges and adopts an organizational style requisite to discussion of them. The paper's goal is to contribute to increasing conceptual clarity and theoretical discernment around the most pressing challenges facing the field of cultural evolutionary science. It will be of most interest to cultural evolutionary scientists, theoreticians, philosophers of science and interdisciplinary researchers.