In 2021 May the Dark Energy Spectroscopic Instrument (DESI) collaboration began a 5 yr spectroscopic redshift survey to produce a detailed map of the evolving three-dimensional structure of the Universe between z = 0 and z ≈ 4. DESI’s principal scientific objectives are to place precise constraints on the equation of state of dark energy, the gravitationally driven growth of large-scale structure, and the sum of the neutrino masses, and to explore the observational signatures of primordial inflation. We present DESI DR1, which consists of all data acquired during the first 13 months of the DESI main survey, as well as a uniform reprocessing of the DESI Survey Validation data, which were previously made public in the DESI Early Data Release. The DR1 main survey includes high-confidence redshifts for 18.7M objects, of which 13.1M are spectroscopically classified as galaxies, 1.6M as quasars, and 4M as stars, making DR1 the largest sample of extragalactic redshifts ever assembled. We summarize the DR1 observations, the spectroscopic data-reduction pipeline and data products, large-scale structure catalogs, value-added catalogs, and describe how to access and interact with the data. In addition to fulfilling its core cosmological objectives with unprecedented precision, we expect DR1 to enable a wide range of transformational astrophysical studies and discoveries.

Precise cosmic web classification of observed galaxies in massive spectroscopic surveys can be either highly uncertain or computationally expensive. As an alternative, we explore a fast Machine Learning-based approach to infer the underlying dark matter tidal cosmic web environment of a galaxy distribution from its β-skeleton graph. We develop and test our methodology using the cosmological magnetohydrodynamic simulation Illustris-TNG at z = 0. We explore three different tree-based machine-learning algorithms to find that a random forest classifier can best use graph-based features to classify a galaxy as belonging to a peak, filament, or sheet as defined by the T-Web classification algorithm. The best match between the galaxies and the dark matter T-Web corresponds to a density field smoothed over scales of 2 Mpc, a threshold over the eigenvalues of the dimensionless tidal tensor of λ _th = 0.0, and galaxy number densities around 8 × 10⁻³ Mpc⁻³. This methodology results on a weighted F1 score of 0.728 and a global accuracy of 74%. More extensive tests that take into account light-cone effects and redshift space distortions are left for future work. We make one of our highest ranking random forest models available on a public repository for future reference and reuse.

This study investigates visual fatigue in a real-world pharmaceutical packaging environment, where operators perform repetitive inspection and packing tasks under frequently suboptimal lighting conditions. A human-centered methodology was adopted, combining adapted self-report questionnaires, high-frequency eye-tracking data collected with Tobii Pro Glasses 3, and lux-level measurements. Key eye-movement metrics—including fixation duration, visit patterns, and pupil diameter—were analyzed within defined work zones (Areas of Interest). To reduce data complexity and uncover latent patterns of visual behavior, Principal Component Analysis was applied. Results revealed a progressive increase in visual fatigue across the workweek and throughout shifts, particularly during night work, and showed a strong association with inadequate lighting. Tasks involving high physical workload under poor illumination emerged as critical risk scenarios. This integrated approach not only confirmed the presence of visual fatigue but also identified high-risk conditions in the workflow, enabling targeted ergonomic interventions. The findings provide a practical framework for improving operator well-being and inspection performance through sensor-based monitoring and environment-specific design enhancements, in alignment with the goals of Industry 5.0.

La valoración del riesgo biomecánico en las actividades industriales es una estrategia que permite la toma de decisiones oportunas en la prevención de aparición de enfermedades osteomusculares en la población trabajadora; sin embargo, los métodos de evaluación tradicionales no permiten hacer una valoración ágil y rápida, siendo procesos que demandan tiempo y experticia del evaluador en ocasiones con resultados sesgados según el criterio del observador. El presente artículo presenta el estado del arte en evaluación de métodos de detección automatizada de riesgo ergonómico en entornos de trabajo industrial, dentro del marco del proyecto de investigación desarrollo de un modelo predictivo de evaluación del nivel de riesgo biomecánico en miembros superiores, en la industria farmacéutica, aplicando técnicas de IA, según metodología OCRA.

This study presents OCRA-AI, a semi-automated computer vision system for biomechanical risk assessment based on the OCRA Checklist methodology. The system integrates three deep learning models — MediaPipe Hands for hand landmark estimation, YOLO-Pose for proximal upper-limb joint detection, and YOLO-grips, a custom-trained model for classifying industrial grip types — and was evaluated in 43 pharmaceutical workers performing repetitive packaging tasks in a real industrial environment.,The YOLO-grips model achieved a mean Average Precision at IoU = 0.5 (mAP@0.5) of 0.804, with per-class performance ranging from 0.671 (left hook grip) to 0.887 (right hook grip). OCRA-AI exhibited laterality-dependent performance. For the non-dominant left upper limb, the system demonstrated moderate concurrent validity with the manual OCRA Checklist (Spearman ρ = 0.50, p < 0.001; ICC(2,1) = 0.34), with near-zero systematic bias in the Bland–Altman analysis (bias = +0.12 ICKL units; MAE = 1.93; MAPE = 11.5%). For the dominant right upper limb, agreement was poor (Spearman ρ = -0.10, p = 0.544; ICC(2,1) = -0.10) despite low absolute errors (MAE = 0.97; MAPE = 5.4%), primarily reflecting the model's convergence toward the population mean. Work shift showed a significant positive association with left-limb risk (ρ = +0.35, p = 0.021), with the highest risk observed during the night shift. Overall, the system reduced assessment time by 93% (28.4 to 2.1 minutes per worker).,OCRA-AI represents a scalable tool for group-level ergonomic surveillance in occupational settings. Individual-level classification, particularly for dominant-limb assessment, still requires expert ergonomist judgment or complementary reference instrumentation.