Progressive deterioration is a problem that affects road infrastructure, especially bridges. This requires the development of methods for its adequate detection and revision, one of them being load testing. Within load testing, finite element analysis (FEA) models provide initial information to understand the behavior of a structure and plan accordingly, which represents a fundamental step towards a precise structural evaluation of a bridge. This study focused on the modeling and analysis of the static response of the bridge over the river Lili in Cali, Colombia, a prestressed girder bridge programmed to undergo a diagnostic load test. A linear FEA model was created with information from a manual survey and from other bridges’ plans designed and built under the regulations in force at the time. Due to the absence of plans and design specifications for the bridge, variations were applied to certain model parameters (stiffness of diaphragms and elastomeric bearings), to quantify their effect on the overall behavior of the bridge. The analysis included obtaining the critical position for the design vehicles, the transversal distribution of stresses and determining the influence of the variation parameters in the response of the structure. Results showed that the critical combinations for bending moment and shear were when the loads were the closest to the exterior girders, being these elements the most affected. The variation on the modulus of elasticity for the diaphragms and the stiffness of the elastomeric bearings did not significantly influence the results for bending moment and shear, nor the critical position. Girder distribution factors (GDF) from the model were compared to previous research, finding similarities in shape and value with other FEA models and experimental results. Finally, an instrumentation plan focused on the girders of the bridge was proposed based on the zones where the maximum effects are expected. The findings in this study show how linear FEA models provide initial but relevant information regarding the critical position of design vehicles, the distribution of stresses and the expected values for bending moment and shear under design loads. ... Read more

Mental workload is an important construct for understanding human cognitive states in conventional, automated, and highly automated vehicles. As vehicle automation progresses, the role of the human changes from active driving to supervision or passive use, which leads to different types of cognitive demands across different levels of automation. Existing approaches for assessing mental workload in automotive contexts have mainly relied on subjective questionnaires and, to a lesser extent, performance-based measures. While these methods have been widely used in conventional and partially automated driving, their suitability for highly automated environments is still under discussion, as the nature of human-vehicle interaction is changing. This workshop focuses on mental workload assessment in conventional, automated, and highly automated vehicles. It brings together researchers and practitioners to critically discuss current measurement approaches and their limitations, and to explore a broader set of assessment methods, including questionnaires, behavioral measures, and neurophysiological techniques. The aim is to create a shared understanding of how mental workload can be assessed across different levels of vehicle automation and to identify directions for improving existing evaluation methods and tools used in this field. ... Read more

Urban heat is a growing concern for public health, energy demand and urban livability, yet the large-scale datasets used to monitor it remain “thin”: they show where a city is hot, but not how heat is experienced on a specific street. For non-expert users such as designers and architects, this makes the data hard to interpret and act on. This project asks how urban heat big data can be thickened with contextual knowledge through visualization, and made meaningful, meaning relevant, trustworthy and interpretable enough to support a judgment.
Following a Research through Design approach structured by the Double Diamond, the project combined a literature review on big data and data thickening with a review of eleven existing platforms and tools. None combined reliable heat data, situated experiential data and a visualization legible to non-experts. Fourteen ethnographic thermal comfort walks in Delft and Rome then captured the contextual factors that shape comfort at street level, and compared them with what existing platforms show. They were grouped into nine recurring themes, including surface heat, vegetation nuances, airflow, refuges and short-lived heat sources, none of which the platforms register. On the Delft route, each factor was further classified by its relation to the big data layer: adding nuance, hotter, cooler or temporary, with a fifth category, opportunity, marking where comfort could be improved.
These findings informed an interactive, map-based prototype that applies three established thickening strategies: onion-structured layering, ground-truthing and error-guided attention. A toggleable UHI effect layer sits beneath color-coded contextual points that are revealed through a hover-lens and open into street-level panoramas with participants’ quotes and photos. The prototype was refined over three evaluation rounds with six participants, including think-aloud sessions and structured user tests.
Non-expert users could explore and interpret the combined data without researcher mediation once the interface explained its own vocabulary, legend and layers. They recognized where their experience matched or contradicted the heat map, and examined the evidence behind each point themselves. An AI-generated interpretation feature, tested with two participants, did not clearly add value beyond its role in building the prototype: one participant doubted its claims and the other found it made the experience passive.
This project applies established big-thick data strategies to urban heat, a domain where, as far as the reviewed literature shows, they have not yet been used, and treats visualization as part of the thickening process itself. The evaluation suggests that thickened urban heat data becomes meaningful to non-experts when the visualization shows where thick data diverges from big data and lets users examine the evidence themselves. With a small sample, these findings are indications rather than validated principles. Future work should test crowdsourced contributions, keep thick data up to date as conditions change, and extend the prototype to other cities and languages.

Digital Prototype — Data Thickening Platform
https://thickening-prototype-3.ai.studio/ ... Read more

This study explores the use of porous disks as a modeling approach for wind turbines in both wind tunnel experiments and computational fluid dynamics (CFD) simulations. Experimental testing is conducted on a 0.2m diameter disk made of wired mesh, measuring its wake for several yaw angles. The experiment is recreated in a CFD environment based on the porous-medium approach. The CFD model is validated against part of the measurements and it is used to further investigate the disk wake. Results indicate the porous disk wake resembles that of a wind turbine, especially at a downstream distance of four diameters, and the CFD model effectively captures the disk behavior. Discrepancies between experiment and CFD appear further downstream mainly due to the wake recovery process being only partially captured by the CFD model. When the disk has a yaw angle the wake is displaced laterally, but less than in a widely-accepted deflection model, and this could be due to the lack of wake curling. ... Read more

This thesis addresses the Minimum-Size Decision Tree (MSDT) problem, which seeks the smallest decision tree that perfectly classifies all training data. Smaller perfect decision trees may generalize better to unseen data, are easier to interpret and validate, require less memory, and have faster inference. These properties are particularly valuable in high-stakes domains such as healthcare (Amann et al. 2020), where decisions must be understandable and auditable, and in resource-constrained systems such as microcontrollers, where memory and computation power are limited. Existing methods for continuous features mainly optimize classification accuracy under a depth constraint, while specialized MSDT methods do not exploit branch-and-bound techniques developed for continuous features and may fail to return a valid perfect tree when the search is interrupted. To address this gap, we extend a branch-and-bound framework for continuous features to the MSDT setting by redefining the objective and adding several problem-specific lower bounds alongside a randomized CART-based upper bound. Additionally, we propose a memory-aware hybrid search strategy to balance efficiency and resource usage. Experimental results show that the proposed approach provides useful anytime performance by maintaining a valid perfect decision tree throughout the search and progressively improving its size as the search continues. Although it proves optimality on fewer difficult instances than a state-of-the-art baseline, it can still return increasingly strong feasible solutions when the search is interrupted. The experiments further show that the pair-size lower bound is the most effective practical bound, while the memory-aware hybrid search strategy reduces the risk of memory exhaustion during search. ... Read more