Comillas Pontifical University

This study examines how Kaizen operational practices contribute to Environmental, Social, and Governance (ESG) performance in manufacturing and proposes an integrative conceptual explanation of this relationship. Moving beyond an efficiency-focused view, Kaizen is conceptualised as a socio-technical system in which routines, people engagement, and learning mechanisms generate sustainability-oriented outcomes. The findings show that Kaizen consistently enhances environmental performance through resource optimisation, waste reduction, and emission control, and strengthens governance via improved transparency, accountability, and participatory decision-making. Social outcomes, however, are more contingent and depend on organisational context and managerial intent. Digital technologies—such as IoT, Big Data Analytics, and Artificial Intelligence—and organisational culture emerge as key mediating mechanisms. Based on qualitative content analysis of semi-structured interviews with experienced Kaizen practitioners, the study proposes a conceptual mechanism linking Kaizen practices to ESG outcomes through the joint mediation of digital transformation and organisational culture, offering both theoretical and managerial contributions.

When simulating three-dimensional flows interacting with deformable and elastic obstacles, current methods often encounter complexities in the governing equations and challenges in numerical implementation. In this work, we introduce a novel numerical formulation for simulating incompressible viscous flows at low Reynolds numbers in the presence of deformable interfaces. Our method employs a vorticity-stream vector formulation that significantly simplifies the fluid solver, transforming it into a set of coupled Poisson problems. The body–fluid interface is modeled using a phase field, allowing for the incorporation of various free-energy models to account for membrane bending and surface tension. In contrast to existing three-dimensional approaches, such as lattice Boltzmann methods or boundary-integral techniques, our formulation is lightweight and grounded in classical fluid mechanics principles, making it implementable with standard finite-difference techniques. We demonstrate the capabilities of our method by simulating the evolution of a single vesicle or droplet in Newtonian Poiseuille and Couette flows under different free-energy models, successfully recovering canonical axisymmetric shapes and stress profiles. Although this work primarily focuses on single-body dynamics in Newtonian suspending fluids, the framework can be extended to include body forces, inertial effects, and viscoelastic media.

This paper proposes a stochastic adaptive robust optimization approach to build the bidding curves of an aggregator managing a fleet of electric vehicles (EVs) participating in the day-ahead and intraday electricity markets. These bidding decisions are made hourly, one day in advance, within an uncertain environment. In this context, uncertainties comprise market prices, as well as driving requirements of EV users. These uncertainties are accounted for by using a set of scenarios and confidence bounds, respectively. In this way, this paper combines classic stochastic optimization techniques with adaptive robust optimization, realistically modeling multiple sources of uncertainty. EVs are equipped with vehicle-to-grid technology so that they can both buy and sell energy to the market. The resulting stochastic adaptive robust optimization problem is solved by using the column-and-constraint generation algorithm, which ensures the attainment of the optimal solution in a finite number of steps. Simulations are run by applying CPLEX under GAMS. A case study demonstrates the effectiveness of the proposed approach. Results show that the bidding decisions of the EV aggregator are sensitive to the uncertainty in driving requirements of EVs, which can be controlled through the uncertainty budget. This highlights the usefulness of the proposed approach to prevent the attainment of suboptimal bidding decisions. Moreover, the good performance of the algorithm in terms of obtaining the optimal solution with computational times lower than 6 min suggests potential for model expansion and increased complexity in future works.

The proliferation of Internet of Things (IoT) devices has greatly enhanced global connectivity but has also amplified cybersecurity risks, particularly from self-propagating malware or black worms. As a countermeasure, some researchers have proposed white worms: benign, self-replicating agents designed to autonomously patch vulnerable systems. Yet, their autonomous behavior raises complex ethical and legal concerns. In this paper, we develop a dynamical model of interacting black and white worms using tools from network epidemiology to explore their co-propagation and emergent behavior across IoT networks. We investigate how parameters related to user response, worm aggressiveness, and network topology shape the system’s stability and their dynamics. Our results show that ethical restrictions, such as reduced autonomy or shorter activity, significantly limit the ability of white worms to suppress botnets. Moreover, network structure plays a decisive role in shaping these outcomes. Overall, the study highlights a fundamental tension between ethical design and practical efficacy: to be truly effective, a white worm must behave in ways that challenge its ethical intent. • Model assesses ethically-constrained white worm effectiveness. • Nonlinear thresholds govern botnet suppression dynamics. • Network topology strongly modulates emergent protection in IoT systems.

El blanqueo de capitales no se realiza en un acto aislado sino en una red de transacciones. Hasta ahora, cada entidad solo veía su parte y no había acceso a la foto completa del proceso. Ahora, la IA puede dar un vuelco en la lucha global contra este delito.

The use of Artificial Intelligence (AI) in healthcare is growing quickly and offers big improvements in medical diagnostics, treatment planning, and patient care. However, people often don't trust AI systems, which prevents them from being widely used. This article looks at both the philosophical and practical issues of trust in healthcare AI systems. First, we provide an overview of the current state of AI in healthcare. Then, we review existing research on trust in technology. Based on our findings, we identify three main factors that affect trust in AI: Technology-Related Factors (transparency, reliability, safety), Healthcare Context Factors (how well AI fits into healthcare settings, proper training for professionals), and Individual User Factors (user experience and attitudes toward AI). Our results show that continuous human oversight, strong regulations, and ethical considerations are essential. Addressing these areas is key to making sure AI systems in healthcare are reliable, transparent, and trusted by both healthcare professionals and patients.

IntroductionBreast cancer (BC) is the most common malignancy among women in Spain and worldwide. Ovarian function suppression (OFS) is recommended as an adjuvant strategy for high-risk hormone receptor (HR)-positive premenopausal BC patients. However, OFS is associated with unfavorable changes in body composition, weight gain, and adverse cardiorespiratory and emotional effects. Multimodal, individualized interventions integrating nutrition, exercise, and psycho-oncological support have demonstrated safety and efficacy in promoting healthy body composition, weight control, and cardiorespiratory fitness (CRF). This study aims to evaluate the impact of a personalized multicomponent intervention in premenopausal BC patients receiving adjuvant OFS.Methods and analysisWe present the ENBI project study protocol proposal as a single-center, open-label, 2:1 randomized pilot study designed to assess the effects of a 12-week individualized program (nutrition, exercise, and psycho-oncological support) versus the World Health Organization (WHO) healthy lifestyle recommendations. Participants are premenopausal HR-positive BC patients undergoing adjuvant OFS. The study duration is estimated at 15 months. The primary endpoint is weight and body composition change, measured via scale and bioelectrical impedance analysis at baseline, post-intervention, and at 6- and 12-month follow-up. Secondary outcomes include CRF, cardiac variability, muscle strength, physical function, laboratory parameters, patient-reported outcomes (quality of life, fatigue, physical activity), systemic therapy-associated adverse events, and nutritional and psychological status. Exploratory outcomes include inflammatory markers (C-reactive protein, tumor necrosis factor alpha, adiponectin) and oncostatin-M. Due to the lack of prior data, the sample size was pragmatically set based on the Hospital General Universitario Gregorio Marañón (HGUGM) Tumor Board registry, with an estimated recruitment of 30 patients.Ethics and disseminationThe HGUGM Drug Research Ethics Committee (CEIm) has approved the study protocol. Results will be presented at national and international conferences and published in a peer-reviewed journal.Trial registration numberNCT06727487

Smart manufacturing demands adaptive, scalable, and human-centric solutions for predictive maintenance. This paper introduces the concept of Agentic AI, a paradigm that extends beyond traditional multi-agent systems and collaborative AI by emphasizing agency: the ability of AI entities to act autonomously, coordinate proactively, and remain accountable under human oversight. Through federated learning, edge computing, and distributed intelligence, the proposed framework enables intentional, goal-oriented monitoring agents to form self-organizing predictive maintenance ecosystems. Validated in a ceramic manufacturing facility, the system achieved 94% predictive accuracy, a 67% reduction in false positives, and a 43% decrease in unplanned downtime. Economic analysis confirmed financial viability with a 1.6-year payback period and a €447,300 NPV over five years. The framework also embeds explainable AI and trust calibration mechanisms, ensuring transparency and safe human–machine collaboration. These results demonstrate that Agentic AI provides both conceptual and practical pathways for transitioning from reactive monitoring to resilient, autonomous, and human-centered industrial intelligence.