http://hdl.handle.net/10498/6783 Wed, 04 Nov 2026 23:06:29 GMT 2026-11-04T23:06:29Z http://hdl.handle.net/10498/40500 RADRISK Inspector: a mobile application for shielding design and safety distance calculation in industrial gamma radiography Suffo Pino, Miguel; Brea-Piñero, D.; Molina Pérez, J. F.; Mota Macías, José Miguel Industrial gamma radiography is a widely used non-destructive testing (NDT) technique, but it involves significant radiological risks that require precise planning of shielding and safety distances. Current practices rely heavily on in-situ radiometric measurements, which are time-consuming and may increase unnecessary operator exposure during setup and verification stages. This work presents RADRISK Inspector (RRI), a cross-platform mobile application that enables real-time radiological safety planning through physics-based calculations. The system estimates minimum safety distances, required shielding thickness, and risk zoning based on radionuclide activity, exposure parameters, and material attenuation properties. In addition, RRI integrates interactive zoning maps, exposure timers, and automated dosimetric reporting compliant with international standards (IAEA, CSN), supporting full digitalization of radiological risk management workflows. The application was preliminarily verified under real industrial conditions during pipeline weld inspections using an Ir-192 source. In this field case, RRI estimated an operator safety distance of 66 m, compared with 22 m obtained from in-situ dose-rate measurements, indicating conservative behaviour consistent with its safety-first planning approach. Accordingly, RRI is not intended to replace in-situ radiometric verification or high-fidelity radiation transport modelling, but to support conservative pre-inspection planning, risk zoning and traceable professional reporting in industrial gamma radiography. RRI represents a novel and beneficial application of radioactivity in modern industry, enabling proactive radiological protection, reducing unnecessary exposure, and improving operational efficiency in gamma radiography. Its compatibility with ATEX-certified devices further supports deployment in hazardous environments. This work supports a shift in radiographic NDT from reactive measurement-based practices to predictive, physics-based safety planning. Thu, 01 Jan 2026 00:00:00 GMT http://hdl.handle.net/10498/40500 2026-01-01T00:00:00Z http://hdl.handle.net/10498/40397 Large Language Models for simplifying directly-follows graphs in support of self-regulated learning Caballero Hernández, Juan Antonio; Palomo Duarte, Manuel; Gutiérrez Madroñal, Lorena; Muñoz Ortega, Andrés; Dodero Beardo, Juan Manuel Directly-Follows Graphs (DFGs) derived from educational process mining provide rich evidence of students' interaction patterns in virtual learning environments, but their complexity often hinders interpretation by instructors and researchers. This study investigates how Large Language Models (LLMs) can support learning analytics by transforming complex DFGs into simplified models and interpretable summaries of student behaviour, with a specific focus on self-regulated learning (SRL). We analyse the actual event log of a compulsory undergraduate course in a Computer Science degree (87 students, ~140k events), discovering a DFG using PM4Py, and compare several GPT-5 variants and reasoning configurations for the task of simplifying this model. The LLM is prompted to remove less relevant transitions while preserving core behaviour. We examine the structural impact of simplification and qualitatively analyse behavioural patterns in a representative simplified DFG. Results reveal a spectrum of trade-offs: smaller models tend to over-prune, whereas larger ones produce more conservative reductions with higher coverage. Despite substantial simplification, the selected GPT-5 configuration preserves navigation, assessment and grade-monitoring patterns that can be interpreted in terms of planning, performance and reflection phases of SRL. We argue that such human-readable process summaries can help instructors detect less effective behaviours, redesign course workflows and scaffold SRL, positioning LLM-assisted DFG simplification as a promising bridge between algorithmic process mining and educational decision-making. Tue, 01 Dec 2026 00:00:00 GMT http://hdl.handle.net/10498/40397 2026-12-01T00:00:00Z http://hdl.handle.net/10498/40370 From Black Box to Glass Box: GenAI in Non-STEM Computing Education Gutiérrez Madroñal, Lorena; Caballero Hernández, Juan Antonio; Baena Pérez, Rubén; Palomo Duarte, Manuel The rapid integration of Generative AI into higher education challenges traditional assessment integrity. This quasi-experimental study analyses its impact on web development skills within an Advertising and Public Relations degree. Drawing on academic performance data, transparency logs, and student perception surveys across four distinct cohorts, the research traces the evolution from a pre-AI baseline through unregulated adoption to a regulated Glass box intervention. Quantitative analysis reveals that while unregulated adoption led to mark inflation, the Glass box protocol stabilised performance. Granular data exposes a dual impact: Generative AI hindered accuracy in foundational tasks due to unverified corrections, yet eliminated resubmissions in complex tasks. The paper contributes evidence-based guidelines for designing non-STEM assessments, demonstrating how Chain-of-Thought documentation converts GenAI from a threat into a verifiable auxiliary tool. Tue, 06 Oct 2026 00:00:00 GMT http://hdl.handle.net/10498/40370 2026-10-06T00:00:00Z http://hdl.handle.net/10498/40330 Methodological pitfalls in predictive mutation testing: threats, impact and open challenges Delgado Pérez, Pedro; Balderas Díaz, Sara; Medina Bulo, María Inmaculada; Guerrero Contreras, Gabriel José Predictive Mutation Testing (PMT) has emerged as a promising technique for reducing the high computational cost of traditional mutation testing. This is achieved by using predictive models to estimate mutant behavior without actually requiring their execution. However, despite its growing interest, PMT is still a young and evolving field and, therefore, its experimental studies remain particularly vulnerable to various methodological threats that may compromise the validity, comparability, and reproducibility of results. This paper aims to identify and examine key methodological threats in prior PMT studies that not only affect their validity, but may also limit the technique’s ability to reach its full predictive potential. For each of the eight identified threats, spanning the main stages of the PMT workflow, we describe its nature, analyze how it has been addressed (or overlooked) in previous work, and offer a validation list of recommended practices. We further conduct an empirical validation to substantiate the impact of these threats. In addition, based on the gaps identified in this analysis, we outline several open challenges that remain unexplored in the field, such as the lack of standardization in dataset sharing, the prevalence of Java-centric studies with method-level operators, and the challenges posed by class imbalance and project heterogeneity. This work contributes to strengthening the methodological soundness of PMT research, promoting more meaningful cross-study comparisons, and encouraging the adoption of practices that foster reproducibility, scalability, and real-world applicability. Thu, 21 May 2026 00:00:00 GMT http://hdl.handle.net/10498/40330 2026-05-21T00:00:00Z