Understanding subjective urban experiences is essential for designing cities that enhance well-being. Urban design should account for the psychological effects of environments on individuals, as these significantly shape perceptions and behaviors. However, a major challenge is the limited availability of urban perception data. Recent studies have leveraged large, crowdsourced datasets like Place Pulse 2.0 (PP2) to inform machine learning (ML) models for urban perception prediction, but the accuracy and reliability of outcomes remain underexplored. There is a critical need to evaluate whether these datasets truly capture human perceptions. This study investigates the role of urban street images in understanding environmental perceptions, using the PP2 dataset and ML techniques. It explores various ML pipelines, employing TPot AutoML for model selection and 5-fold cross-validation to prevent overfitting. The goal is to identify the most efficient model that strengthens the link between automated predictions and human perception. The study also applies SHAP (SHapley Additive exPlanations) to interpret model outputs, revealing feature importance and interactions. This improves transparency and ensures ML-generated insights are actionable for urban planning. By rigorously testing ML pipelines, this research enhances predictive accuracy and contributes to the development of reliable urban design tools. The findings highlight ML’s potential in processing large-scale perception data, uncovering hidden patterns, and informing people-centered urban planning. However, further validation against real-world surveys is necessary to ensure robustness and generalizability in assessing urban perceptions.
Strategic Classification of the Integration Between Artificial Intelligence (AI) and Building Information Modelling (BIM): Opportunities and Future Challenges
The integration of Artificial Intelligence (AI) and Building Information Modelling (BIM) represents a promising but complex frontier in the construction industry. While BIM has already transformed construction workflows through digitalization and lifecycle management, AI has the potential to further advance automation, optimization, and data-driven decision-making. This study aims to provide a systematic classification of AI applications within BIM, identifying four key categories: AI as a digital consultant, collaborative interface, process regulator, and process outcome. Through a scientometric analysis of literature and structured mapping of existing software, the paper evaluates the state of the art, exploring the capabilities and limitations of current solutions. Results highlight how AI + BIM tools are transforming various lifecycle stages, from conceptual design to construction and operations. However, challenges remain, including the lack of standardization, risks related to data security, and the balance between automation and human oversight. This classification framework not only structures existing knowledge but also directs future research and applications, encouraging critical reflection on AI’s role in advancing BIM methodologies while considering its implications for transparency, efficiency, and technological governance.
Artificial Intelligence and Extended Reality for Communicating the Uses of Natural Fibers in Building Construction: State of the Art and New Proposals
This contribution presents a digital framework to promote and communicate the use of natural fibers in building construction, developed within the “Circular Design for Natural Fibers” (CD4NF) project. The research addresses the limited application of advanced digital tools in this emerging field by proposing an integrated workflow that synergizes Building Information Modeling (BIM), Artificial Intelligence (AI), and eXtended Reality (XR). The core of the proposal is a flexible and modular platform built upon a structured material filing system that catalogs detailed information on various natural fibers. This platform utilizes BIM as a central data repository. An AI system, based on a Retrieval-Augmented Generation (RAG) model, enables users to query this complex data using natural language. XR applications provide an interface for visualizing and interacting with the BIM models and their associated information in real-world contexts. This approach aims to optimize key processes from material research and design to construction, enhancing decision-making, interoperability, and communication. The system represents a step towards a more efficient, circular, and sustainable construction sector by facilitating the informed adoption of natural fiber-based materials.
The Language of Photography in the Age of AI
Text-to-image algorithms based on Deep Learning are central to content creation in multiple application domains. In the last few years, the capacity of Neural Networks to generate increasingly realistic images quickly has blurred the boundary between authentic and realistic content, making genuine and false data less and less distinguishable. This condition leads to a profound reflection on the application of photographic images as a tool for communication and storytelling, trying to answer simple questions. Can today’s Neural Networks generate content comparable and indistinguishable from a photograph in both formal and compositional terms? Can artificial intelligence algorithms replace the photographer’s ability to design and obtain images that preserve the story and the place’s intangible culture? From a set of photographic rules framed in specific workflows, the research analyses some results obtained using text-to-image algorithms within the Midjourney program. The experiment aims to determine the pros and cons of using text-to-image algorithms to automatically generate photographic images, highlighting the potential and current limitations in constructing content subject to specific formal rules.
Exploring Evolutionary Optimization: Integration of AI and Additive Manufacturing
Contemporary architecture, as well as design, has revolutionised the approach to form creation, prioritising increasingly efficient and, above all, adaptive modelling. It is essential for designers to identify the requirements that the product to be realised must meet for its own production. This methodology, driven by technological advancements in the field of artificial intelligence (AI) and the use of advanced algorithms, enables the exploration and generation of optimised products from various perspectives based on chosen criteria; the resulting forms integrate functionality and aesthetics. This research proposes the use of tools within the parametric modelling environment Grasshopper, such as Galapagos and Octopus, which employ AI algorithms to enhance the design process and optimise forms for additive manufacturing. Experimentation with these plugins allows leveraging evolutionary algorithms to explore a wide range of design solutions, enabling designers to efficiently optimise complex forms. In this context, AI facilitates tackling multi-objective optimisation problems, improving parameters such as structural strength, material usage, and minimisation of printing times. This approach not only enhances the efficiency of the design process but also opens up new possibilities for innovation in design by integrating the advanced computational capabilities of AI with the creative potential of parametric design.
The Use of the Imagematching Software and Other AI Tools for the Automatic Recognition of Similar Images: Some Theoretical Considerations
The article discusses the new digital research methodologies to investigate the concept of image similarity, taking as a case study the Lyon16ci database project. Developed in collaboration with the Visual Geometry Group at the University of Oxford, this project investigates how AI-driven image recognition can enhance scholarly analysis of visual material in the humanities. The focus is on the use of VISE software, designed to automatically retrieve visually similar images based on geometric and compositional features. The article provides a critical evaluation of the strengths and limitations of this tool in the context of art historical and visual culture research. It discusses how VISE facilitates new interpretative approaches by uncovering visual relationships and how it can effectively enhance traditional comparative methods. The author provides an overview of the advantages and constraints of using the VISE AI software to automatically retrieve similar images, presenting some of the theoretical considerations and the research possibilities provided by image recognition tools. The Lyon16ci case offers insights into the broader potential of machine vision in redefining the scope and scale of such image-based humanities research.
Revealing and Interpreting Complex Urban Patterns from Location Based Social Network Data. An Investigation into Chinese Stadiums in the Global South
This paper focuses on the adoption of big data visual representation and semantic interpretation to study complex urban patterns in places largely impervious to traditional mapping technologies and documentary analytical tools. In particular, it examines stadiums built by China in the Global South as part of a broader strategy of building diplomacy. Between 1959 and 2022, China facilitated approximately 2,000 construction projects in developing countries, among which more than 150 are large-scale sports facilities. While these buildings have mainly been analyzed from architectural and typological perspectives, their relationships with the surrounding urban environments and their effects on local communities and ecosystems have remained largely unexplored and difficult to interpret. This paper demonstrates how big data — particularly Location-Based Social Network (LBSN) data — together with visualization and AI-supported interpretation systems, can provide new opportunities to understand the capacity of these large-scale architectural infrastructures to attract people, influence movement patterns throughout urban space, and generate economic and social impacts on the existing city.
Representation Across Boundaries: New Paradigms in the Age of AI and XR
The introduction and rapid expansion of new algorithms based on Machine Learning (ML) and Deep Learning (DL) processes to support knowledge and design activities has revolutionized multiple domains in recent years. Among these, research in the fields of Cultural Heritage, Design, and Architecture is fostering the development of new methodologies for study and content creation — partly supporting existing tools and partly replacing them entirely — while offering a new paradigmatic perspective on the impact of AI within these domains.
More specifically, the introduction of Generative AI (GenAI) and the creation of new forms of content within these fields open new possibilities for the understanding, analysis, design, and communication of architecture and design. At the same time, these developments highlight the limitations and risks associated with their uncritical use and raise important ethical questions. Human guidance and supervision in generative processes still remain — fortunately — a foundational component of these workflows, ensuring control over results while encouraging their implementation across different areas.
Through a concise review of current research in the field, the article provides an updated overview of recent international studies, while anticipating possible future developments concerning XR and AI in Cultural Heritage, Design, and Architecture.
Is a Picture Worth a Thousand Words? Comparative Evaluation of Generative AI for Drawing and Representation
The paper presents a comparative evaluation of generative AI systems for drawing and architectural representation, focusing on the operational principles, workflows, and visual outputs of text-to-image applications such as Midjourney, Stable Diffusion, and DALL-E 2. The research analyzes neural networks, latent space mechanisms, generative adversarial networks, autoregressive models, and diffusion probabilistic models to explain how AI systems transform textual prompts into visual representations. Through experimental prompt engineering and comparative image generation tests, the study investigates the relationship between AI-assisted creativity, visual storytelling, representation processes, and design workflows. The paper critically discusses the implications of generative AI for architecture and visual culture, including authorship, ethics, copyright, bias, realism, and the transformation of creative practices. The research ultimately proposes an informed and critical approach to AI-assisted representation, emphasizing the evolving role of designers as curators and strategic decision-makers within AI-driven creative environments.
Hypotheses of Images and Architectural Spaces in the Age of Artificial Intelligence
The paper explores the relationship between artificial intelligence, architectural representation, and digital techno-cultures through experimental workflows combining text-to-image generation, text-to-3D modeling, parametric design, and AI-assisted visualization. The research investigates how generative AI tools such as Midjourney, ChatGPT, PointE, Dreamfusion, and Grasshopper can support the creation of architectural forms, semantic image transitions, authorial hybridizations, and morphogenetic spatial configurations. The study proposes a conceptual framework organized around different AI media categories, including textual AI, image generation, video generation, post-production systems, and AI-assisted parametric modeling. Through hybrid workflows integrating AI-generated Python scripts, parametric modeling in Rhinoceros/Grasshopper, and generative visual experimentation, the paper reflects on the epistemological, creative, ethical, and technological implications of AI in architecture and representation. The research emphasizes AI as a creative and assistive medium capable of generating new spatial hypotheses and experimental design processes within architecture and digital representation.