Advances on Design Engineering V

Table of Contents

1. Frontmatter

2. Digital Tools and Artificial Intelligence

   1. Frontmatter

   2. AutoCAD Automation Through Python Scripting

      Antonio Fernández, Lucía Díaz-Vilariño, Francesco Bianconi

      Abstract

      This work explores the automation of AutoCAD workflow using Python scripting, with a focus on the different Python libraries available for such tasks. The article analyzes four prominent Python packages—win32com, pyautocad, PyAutoGUI, and ezdxf—assessing their strengths and weaknesses in the context of AutoCAD automation. To evaluate the performance of these libraries, a Python application was developed for the simple task of prefixing layer names in an AutoCAD drawing, testing each library’s effectiveness in accomplishing this task. The findings indicate that each Python library offers distinct advantages depending on the automation requirements, with ezdxf standing out in aspects such as versatility, robustness, and documentation quality.

   3. Reframing Urban Morphology for AI: Integrating Qualitative Dataset for a Specialized Artificial Intelligence

      Caterina Juric, Ezgi Nur Güngör, Alessandro Lovisolo

      Abstract

      To explore the potential of a customized GPT-4o language model in under-standing and analysing urban morphology theory, three influential books were carefully selected: The Image of the City by Kevin Lynch (1960), Townscape by Gordon Cullen (1961), and The Death and Life of Great American Cities by Jane Jacobs (1961). These books, cornerstones of urban form studies, were chosen for their shared historical context, thematic alignment, and methodological complementarity. Starting from three prompts, the Concept Correlator AI was taught to describe the books through keywords and tables and connect the different words through diagrams. By leveraging this glossary, the model analysed prompts related to these concepts, drawing from a comprehensive collection of sources to provide well-informed, concise responses. It systematically compared definitions across the books uploaded in its folders, grouping similar ones under a unified category. Through this refined glossary, the study evaluated AI's ability to interpret urban forms concepts and identified gaps between the existing definitions and the responses of the customized GPT-4o model. Ultimately, this framework strengthened AI's foundational knowledge, improving both the contextual accuracy of responses and computational efficiency in urban morphology research. Some definitions in the analysis were identified as incorrect and required reformulation of the lexicon by the AI. This targeted data organization enhances research efficiency by minimizing unnecessary data processing and improving the precision of AI-generated outputs. By bridging traditional typo-morphological theories with emerging AI methodologies, the study contributes to AI's systematic and meaningful integration into urban morphology, advancing theoretical understanding and practical applications.

   4. Design and Development of an Affordable, Ergonomic Bionic Hand for Improved Daily Functionality

      Isabel Segui Verdú, Alba Rey De Viñas Redondo, Larisa Dunai, Luis Tarocher

      Abstract

      A hand loss or severe impairment can significantly impact an individual’s quality of life by restricting essential daily activities and professional tasks. Traditional prosthetic solutions face challenges regarding cost, accessibility, and functionality, limiting their widespread adoption. This study presents an innovative approach to designing and developing a bionic hand prosthesis that integrates ergonomic principles, intuitive control, and sustainable manufacturing methods. The primary focus is affordability and usability, offering an accessible and adaptable alternative to conventional prostheses. The design process leverages CAD tools for precise modelling and simulation, ensuring structural accuracy and efficiency. Additive manufacturing uses biocompatible, cost-effective, and eco-friendly materials to balance durability, robustness, and comfort. The system enables precise control by integrating sensors such as inertial measurement units (IMUs), enhancing user experience and adaptability. The control system processes inputs to facilitate intuitive operation. Eliminating reliance on EMG expands the prototype’s applicability to a broader range of users, including those with minimal residual muscle activity. With its affordable and sustainable design, this prosthesis has the potential to provide access to functional prosthetic solutions, particularly in developing countries, ultimately improving the quality of life for individuals with upper limb disabilities.

   5. Augmented Reality for Machinery Layout Optimisation in Olive Oil Mills: An Innovative Approach

      Diego Francisco García-Molina, Jorge Manuel Mercado-Colmenero, Adonis Jose Pabuena-García, Cristina Martín-Doñate

      Abstract

      In industrial environments, the organisation of space and layout is crucial to achieve efficiency in production processes. In the olive oil industry, traditional methods are based on CAD tools and two-dimensional representations, leading to problems of spatial interpretation and an increase in installation errors. The incorporation of Augmented Reality (AR) offers a more innovative way to interact, improve planning accuracy and avoid operational costs by representing 3D models in real environments. In this study, an application is created that uses AR for planning the design of olive oil extraction machines in mills. The interactive spatial planning methodologies applied in the system consider relevant parameters including, among others, free areas, operational safety distances and workflow efficiency. The system was created using Unity 2022. The application enables users to locate, move and zoom machine drawings in real time, which considerably increases the accuracy of the design prior to installation. A study was carried out on the impact of augmented reality on machinery distribution, in contrast to traditional planning methods in the oil industry. A feasibility study was also carried out to assess the application, which included the integration of 3D models of the machinery in an augmented reality environment. Augmented Reality-based simulations were developed to allow operators and technical staff to visualise and interact in real time with the proposed equipment chains. The results emphasise the significant contribution of AR to optimise spatial planning and decision making in the olive oil industry. AR reduces machinery relocation costs and improves overall efficiency by making it easier for operators to identify installation errors before any physical modifications are made. Furthermore, AR democratises in-industrial design through an intuitive method, enabling planning and design by professionals who are not experienced in CAD software. Unlike commercial CAD tools that often require extensive technical expertise, this application allows users to interactively explore virtual models in the physical working environment, bridging the gap between design and final implementation.

   6. Custom Geometrical Reconstruction of Human Corneal Surfaces Based on NURBS Curve Optimization by An Evolutionary Algorithm: A Preliminary Study

      Francisco L. Sáez-Gutiérrez, José S. Velázquez Blázquez, Francisco Cavas Martínez

      Abstract

      Recent advances in surface reconstruction from 3D point cloud data have significantly improved available techniques used for generating accurate digital models of physical objects. Among these, new methods have been proposed for generating Non-Uniform Rational B-Splines (NURBS) surfaces from 3D imaging data using machine learning algorithms. This study introduces a novel approach to the three-dimensional modal reconstruction of the human corneal surface using an evolutionary algorithm-based method. The proposed workflow employs an evolutionary algorithm to optimize the zonal reconstruction of the anterior and posterior corneal surfaces, ensuring precise alignment and smoothness of the resulting geometry. Genetic algorithms are particularly effective for optimizing NURBS curves with Galapagos software, making them well-suited for tasks such as minimizing approximation errors relative to reference points, enhancing curve smoothness, or achieving specific configurations in applications like parametric design, engineering, or manufacturing. The process begins with an initial population of NURBS curves, which are generated with randomized configurations of control points, weights, and parameters. To enhance convergence, curves can be based on an initial baseline design. All data used in this study was obtained from the IBERIA BIOBANK database (Universidad Miguel Hernández de Elche, OFTARED–Instituto de Salud Carlos III). The spatial point cloud data was integrated into a parametric workflow using Grasshopper software. The inclusion of an evolutionary algorithm introduces a significant layer of optimization to generate NURBS surfaces. This method adapts to the topographic data, iteratively refining the corneal model based on fitness criteria. The evolutionary algorithm operates over multiple generations, evaluating, selecting, and reproducing new NURBS curves. The process continues until a stopping criterion is reached, such as a maximum number of generations or a predefined error threshold. This study presents a novel framework for corneal modelling using an evolutionary algorithm-based method integrated with Galapagos software. By combining parametric design and advanced optimization techniques, this methodology establishes a foundation for innovative diagnostic applications in ophthalmology. The potential benefits include improved understanding and management of corneal diseases, contributing to advancements in clinical practices.

   7. Exploring Large Language Models for Cad Automation: A Case Study with CATIA Scripting

      Héctor de Pablo Pascual, Santiago Delgado Vaquero, David Escudero-Mancebo

      Abstract

      Objectives: Generative AI has emerged as a mature technology poised to enhance efficiency in various fields traditionally considered exclusive to human expertise. Computer-Aided Design (CAD) is one such domain that could benefit from this technological evolution. This study explores the potential of Large Language Models (LLMs) in generating code for creating 3D models through text prompts, specifically in CATIA using its scripting module. Materials and Methods: A series of modeling challenges were defined with increasing levels of complexity. LLMs were prompted to generate scripting code capable of constructing corresponding 3D models in CATIA. The generated scripts were tested, and their outputs were analyzed based on accuracy, completeness, and the need for manual intervention. Results: The experiments revealed that while LLMs can generate functional code for 3D modeling tasks, manual review and refinement are often required to ensure correctness and usability. The study also highlighted limitations in model comprehension of geometric constraints and parametric design principles. Conclusions: This research highlights the potential of integrating reinforcement learning techniques into CAD scripting. By controlling 3D modeling through scripts it is possible to design automated processes that adjust parametric models dynamically. Future work could explore how AI-driven optimization strategies can enhance geometric modeling and automate complex design decisions.

   8. Artificial Intelligence in Museums: An Overview of Facial Expression Analysis Techniques in Museum-Space-Visitor Interaction

      Giusi Castaldo, Elena Laudante, Elidia Beatriz Blázquez-Parra, Enrique Dominguez Merino, Mario Buono

      Abstract

      With the rapid development of new technologies, recent advances in Artificial Intelligence (AI) are promoting the growth of the Cultural Heritage field and changing approaches to the design of museum spaces. The museum visitor experience is evolving towards greater user involvement, creating new dynamic and interactive cultural contexts. The design of museum spaces and, in particular, the organisation of visual, spatial and material elements are more oriented toward the needs and expectations of visitors, thus generating an impact on direct involvement and the quality of the visit. In this context, among the tools for capturing user engagement data, facial expression analysis enables the collection of implicit and explicit measures of user activation and emotional engagement in application fields ranging from neuromarketing to user experience. To support this analysis, Artificial Intelligence technology plays a key role through its ability to generate data and algorithms to “reshape” the creative process beyond traditional design methods and tools derived from the study of human perception. This paper proposes a review of the state of the art of the main digital technologies used in museum exhibitions and the identification and detailed study of integrated systems and two methods used for facial emotion recognition during the enjoyment of museum works and spaces through examples and case studies in the literature. Also, the study highlights the importance of facial recognition technology, capable of identifying and classifying different visitor reactions, among the metrics for data collection in the museum field already found in the literature. Moreover, this technology, used in the Cultural Heritage field, is identified as a possible key factor in the construction of guidelines for designing museum spaces in line with current research areas related to the advanced visitor experience in the enjoyment of cultural heritage.

   9. Integration of Artificial Intelligence Tools in Architectural Design and Construction Planning Processes. Case Study of a Single-Family House in Ecuador

      Byron Andrés Morales-Chamorro, Eduardo Vázquez-López

      Abstract

      The integration of artificial intelligence within the construction sector is effecting a paradigm shift in design and planning methodologies. In the realm of recent innovations, large language models (LLMs) have garnered significant attention. The effective integration of LLM with building design remains challenging, particularly in contexts that require adaptation to stringent local regulations. The objective of this study is to analyse the potential of LLM to support building design and construction planning. The study explores the impact of general-purpose LLM on process automation, resource optimisation, and regulatory compliance. The methodological approach adopted in this study is centered on a basic building design in Ecuador. The designed case study will be tested through specific prompts into LLM tools to validate its compliance with local construction regulations. Furthermore, LLM tools are tested in order to generate a preliminary time and cost plan. A comparison of experiment results with real data is undertaken for the purpose of evaluating the efficacy of the predictions made by LLM. The findings of the experiment demonstrate that LLM tools exhibit superior performance in the domain of regulatory compliance certification design in comparison to their efficacy in cost and time planning. Nonetheless, it is imperative that a more extensive array of tests be conducted to ascertain the reliability of LLM tools. In any event, the accelerated development of general-purpose LLM has the potential to enhance efficiency and address the existing unreliability.

   10. Digital Twins Models Applied to CXL Treatments of Corneas Affected with Keratoconus

       José González-Cabrero, Carmelo Gómez, Francisco Cavas

       Abstract

       Keratoconus is a structural cornea disorder due to a combination of genetic, environmental and hormonal variables that leads toa gradual thinning and outward protrusion of the cornea that in initial phases can be treated with cross-linking (CXL) techniques to make the cornea stronger. CXL involves placing riboflavin on the corneal surface and then applying UV-A light to enhance the bonding between collagen fibers within the cornea. The analysis of treatment performance over time can be carried out using personalized digital corneal models that incorporate the mechanical alterations induced by CXL. For this purpose, a variation of the parameters that define the stiffness contribution in hyperelastic materials can be considered. The development of computational corneal models opens the possibility to analyze the local effect of CXL techniques with the development of sectorized in-silico patient specific model where different materials can be considered in different zones optimizing the riboflavin use and the effectiveness of the treatment. The possibility to vary the stiffness in the thickness can also contribute to simulate with more accuracy the real effects of CXL techniques. Computational models can contribute to create a digital twin of corneas with CXL treatment facilitating follow-up and evolution after treatment of these ones.

   11. Assessing the Current Use of Isolated Buildings in Andalusia Through AI Techniques

       Cristina Torrecillas, Francisco J. Ramos-Sánchez, Natalia Martinez-Gómez

       Abstract

       The Andalusian Geographical Gazetteer is a toponymic geodatabase with over 245,000 geographical place names and more than 320,000 locations. In recent years, a study revealed that many isolated buildings in rural environments have experienced a degradation in their use. Specifically, cases were identified where buildings have become dilapidated due to abandonment, completely disappeared, or been demolished and replaced by new constructions, which may also have different names. This study focuses on a sample of 285 entities classified as Rural Buildings, either active or disappeared/in ruins in three municipalities. Alphanumeric files, two different aerial orthophotos imagery (including both the most current and historical images), and a cadastral dataset were used as input data. The methodology was developed within a GIS and Python environment, employing three different methodologies: two artificial intelligence-based methods (Large Language Model and pre-trained Deep Learning Models for Building Footprint Extraction) with images, and a GIS geoprocessing over the Spanish Cadastre. The temporal change of the buildings was classified and studied. The combined results of the three methodologies showed success rates exceeding 92.9%, reaching 100% in one of the three municipality. These figures encourage further exploration of this novel methodology, particularly the use of LLMs and the combined application of various methods to reduce false positives compared to individualized and manual techniques.

   12. Integration of Artificial Intelligence and Open-Source Tools for Intelligent Natural Language Queries on IFC Models: An Accessible and Collaborative Solution

       Norena Martín-Dorta, Ana Pérez-García, Ángel Díaz-Murillo

       Abstract

       To maximize the potential of Building Information Modelling (BIM) projects and to ensure effective collaboration between multidisciplinary teams, efficient information management is essential. This paper presents a solution integrating open source openBIM tools with advanced artificial intelligence to perform intelligent queries on IFC model data. The solution is implemented as a plugin for Blender and Bonsai (formerly known as BlenderBIM) and utilizes the IfcOpenShell library to process and extract model data, including class attributes and property sets. The data is organized in a structured JSON file to facilitate user queries. The proposed workflow begins with users formulating natural language queries in a text box within Blender. These queries, along with the extracted model data, are processed by an advanced AI system, such as Gemini, which can interpret natural language and analyze complex data. The AI retrieves the relevant GUIDs that match the specified criteria, enabling the application to visually highlight the corresponding elements of the BIM model in the 3D environment, changing their colour for easy identification. This solution emphasizes the use of open-source tools, ensuring accessibility, adaptability, and flexibility for both developers and end users. By utilizing Blender, Bonsai, and IfcOpenShell, the workflow promotes interoperability and allows the creation of custom applications tailored to the specific needs of the construction industry. Additionally, the natural language query capability makes BIM model information more accessible, enabling both technical and non-technical users to interact with models intuitively. This enhances decision-making efficiency and fosters interdisciplinary collaboration.

   13. Innovation in Personalized Ambient Assisted Living (AAL) Technologies for Older Adults: From Wearable Devices to Integrated Care Ecosystems

       Eva Tausiet, Álvaro Marco, Marta Siguín, Roberto Casas, Teresa Blanco

       Abstract

       The increase in population life expectancy poses growing challenges in terms of care, autonomy, and quality of life. Although wearable technologies have demonstrated significant potential in health and social care, many current solutions remain closed, rigid systems that poorly accommodate the functional and technological diversity of older users. This paper presents the design of a product-service concept built upon a modular technological ecosystem that integrates multiple wearable devices, fixed sensors, hybrid connectivity, and a control application. The system is conceived to accommodate a wide range of user profiles in both domestic and community settings. The system architecture follows the IoT Thing–Cloud–App model, structured using the Cosica methodology, and enables the progressive integration of devices according to user needs. The primary wearable acts as the central node, managing connectivity contextually via Bluetooth or 4G. An application—envisioned in differentiated versions for users and caregivers—provides an accessible platform for supervision and configuration, aiming to reduce uncertainty and preserve the perception of control. The cloud infrastructure supports data synchronisation and remote services, and is being designed to enable the future integration of AI-based analytics for early detection and intervention. This proposal addresses key challenges at the intersection of ageing and technology, including variations in technological literacy and cognitive capacity—towards more inclusive, flexible, and user-aligned technologies. Beyond the system concept, the work seeks to advance methodological approaches to the design of technological ecosystems and offers insights into relevant design considerations for ageing populations.

   14. Spatial Data Analysis Towards Achieving Artificial Access Consciousness Using Knowledge Graphs, Large Language Models and Graph-Driven Reasoning

       Javier Arévalo-Royo, Juan-Ignacio Latorre-Biel, Francisco-Javier Flor-Montalvo, Eduardo Martínez Cámara

       Abstract

       Artificial access consciousness is an open ambition in contemporary artificial intelligence, concerned with designing systems capable of perceiving, processing and responding to spatial information with a degree of contextual sensitivity. This study examines how spatial data analysis, when combined with knowledge graphs, large language models and graph-based reasoning, might serve as a conceptual pathway toward such a goal. Rather than offering practical results, the present work is structured as a theoretical exploration. It proposes that the integration of generative and inferential capabilities from language models, together with the formal structure of knowledge graphs, could diminish interpretive errors and anchor reasoning more firmly in context. Spatial data analysis remains central to domains such as engineering, construction and agriculture. In these settings, artificial access consciousness, if realized, would allow systems not only to optimize resource allocation or improve site monitoring, but also to align interventions with the real complexity of geographical and environmental variables. The frameworks discussed here are advanced as a foundation for further inquiry, with the recognition that genuine artificial access consciousness will require a sustained and multidisciplinary effort, bridging technical innovation with nuanced interpretation.

   15. Artificial Intelligence as a Tool for Translating Abstract Languages into Architecture

       Cintya Eva Sánchez Morales, José Carlos López Cervantes

       Abstract

       Throughout history, architectural design has relied on human intuition and both manual and digital representation tools to develop spatial concepts. The emergence of generative AI has introduced a new paradigm, allowing architects to integrate transdisciplinary inputs into their design processes in unprecedented ways. Rather than replacing human creativity, AI expands the architect’s conceptual framework, broadening the scope of design possibilities. This paper explores how generative artificial intelligence (AI) enables the transformation of abstract visual languages, particularly Suprematist paintings, into architectural form. Drawing from historical and contemporary examples, the study situates AI within a broader lineage of transdisciplinary design practices. We propose a methodology for translating key compositional principles from Suprematist artworks into spatial configurations. Rather than re-placing human creativity, AI serves as a tool for expanding the architect’s conceptual framework, offering new pathways for experimentation, iteration, and transdisciplinary translation. This methodological approach aligns with Lyotard’s notion that translating sensory structures across mediums preserves traces of their origin, fostering transformation rather than mere replication.

   16. AI for Automated Processes in CAD: An Overview

       Eva Rodríguez Hortelano, Maider Iturrondobeitia Ellacuria

       Abstract

       According to the RAE (Royal Academy of the Spanish Language), Artificial Intelligence (AI) is defined as follows: “Part of computer science that studies and develops systems that imitate human intelligence to perform tasks that naturally require it, such as natural language processing, pattern recognition or machine learning.” Nowadays the artificial intelligence has changed the game, and specifically the education world. In addition, the different applications of the new technologies have change dramatically the classroom aspect, from the blackboard to the digital classroom and from the physical presence to the digital classrooms. The graphic design teachers need to be aware of the innovative AI trends. On the one hand, because the students are asking for them, and on the other hand because there are serious advantages in the CAD software that include AI technology. A groundbreaking current trend consist in AI integrated into the CAD software, in order to improve and automate various aspects of Assisted Design. Here, a review of both fields, CAD and AI, is presented. There is a description of the knowledge area, and developments of both fields over time are depicted and compared. Finally, the advantages of AI application in CAD is analyzed. Those include the improvements of typical software, such as design automation, a performance optimization, a design assistance, a predictive analytics, an advanced simulation and a smart interface. Besides, there is a bigger improvement in the field; there is a new flashy software that is known as Generative Design. This new field is also defined and analyzed, purposing new Generative Design software for the CAD lessons and placing them in the engineering students PBL process.

   17. Identification of Biofouling on Submerged Surfaces: Image Analysis for Deep Learning-Based Approaches

       Emmanuele Barberi, Maria Francesca Alberghina, Luciana Randazzo, Michela Ricca, Felice Sfravara

       Abstract

       Biofouling, the phenomenon involving the growth and accumulation of marine organisms on submerged surfaces, poses a significant challenge for the conservation of underwater structures and historical artifacts. Its presence can accelerate material degradation and complicate potential restoration efforts, especially in cases where direct recovery from the marine environment is not feasible. In this context, the early identification of biofouling is crucial for developing targeted intervention strategies and minimizing long-term damage. This study aims to lay the groundwork for creating a training dataset for a deep learning-based segmentation network. Images of submerged specimens were captured in both the visible spectrum and under UV illumination, which reveals biofouling often invisible to the naked eye. Using semi-automatic thresholding methods, segmentation masks were generated from UV images to approximate biofouling regions. RGB, HSV, Lab, and YCbCr color spaces were systematically compared to determine the most stable and reliable color model for segmentation. Results identified the HSV space as offering the most consistent thresholding performance across the dataset. These segmentation masks provide a foundation for training deep learning models aimed at automatically detecting biofouling in visible-light images, where manual annotation is challenging.

   18. Modeling and Digital Fabrication of the F50 SailGP Catamaran Using Catia V5, FDM Technology and Deep Learning Tools

       José Serrano Gómez, Manuel Morato Moreno, Iker Rodríguez Vega

       Abstract

       The F50 SailGP catamaran is one of the most advanced vessels in competitive sailing. This study presents a multidisciplinary approach combining CAD modeling, additive manufacturing, and deep learning for the digital reconstruction and recognition of the F50. Using CATIA V5, a detailed virtual model was developed based on publicly available visual data, followed by the fabrication of a scaled prototype using FDM technology. The CAD models were segmented for printability, and post-processing ensured dimensional accuracy and structural integrity. In parallel, a YOLO-based computer vision model was trained on a manually labeled dataset of 100 images to identify the F50 in diverse visual contexts. The model achieved 83% precision, demonstrating the feasibility of automated recognition despite limited data. This work highlights the potential of integrating engineering design, rapid prototyping, and AI tools in the analysis and dissemination of high-performance sailing technologies, offering a foundation for future developments in configuration-specific recognition and automated design validation.