Purpose: This study aims to present a twofold machine learning (ML) model, namely, EDU-AI, and its implementation in educational buildings. The specific focus is on classroom layout design, which is investigated regarding implementation of ML in the early phases of design. Design/methodology/approach: This study introduces the framework of the EDU-AI, which adopts generative adversarial networks (GAN) architecture and Pix2Pix method. The processes of data collection, data set preparation, training, validation and evaluation for the proposed model are presented. The ML model is trained over two coupled data sets of classroom layouts extracted from a typical school project database of the Ministry of National Education of the Republic of Turkey and validated with foreign classroom boundaries. The generated classroom layouts are objectively evaluated through the structural similarity method (SSIM). Findings: The implementation of EDU-AI generates classroom layouts despite the use of a small data set. Objective evaluations show that EDU-AI can provide satisfactory outputs for given classroom boundaries regardless of shape complexity (reserved for validation and newly synthesized). Originality/value: EDU-AI specifically contributes to the automation of classroom layout generation using ML-based algorithms. EDU-AI’s two-step framework enables the generation of zoning for any given classroom boundary and furnishing for the previously generated zone. EDU-AI can also be used in the early design phase of school projects in other countries. It can be adapted to the architectural typologies involving footprint, zoning and furnishing relations.

This study investigates how artificial intelligence (AI) can be integrated into first-year design education. The goal of the study is to examine how students combine orthographic projections with AI-generated images. To answer this question, a design representation course called Visual Communication I (VC-I), taught at Istanbul Technical University’s Faculty of Architecture, was selected. A three-step assignment called “ISO-meets-AI” was planned and implemented in the VC-I. The steps of the assignment were as follows: (i) producing orthographic projections of created physical compositions; (ii) generating image output from text input using an AI program (Midjourney); and (iii) combining the AI-generated images with the orthographic projections. The assignment was completed by 50 students from the departments of architecture, interior architecture, and industrial design. Tutors defined the evaluation categories and sub-criteria based on an initial analysis. While tutors evaluated every submission based on the established categories and sub-criteria, students only assessed their own work through a survey. Lastly, a second survey was given to the students in order to better understand their willingness to employ AI in their future studies. The first survey’s comparative evaluation results of the tutors and students, as well as the results of the second survey, are presented.

This study sheds light on a holistic understanding of muqarnas with its historical, philosophical and conceptual backgrounds on one hand and formal, structural and algorithmic principles on the other hand. The vault-like Islamic architectural element, muqarnas, is generally considered to be a non-structural decorative element. Various compositional approaches have been proposed to reveal the inner logic of these complex geometric elements. Each of these approaches uses different techniques such as measuring, unit-based decoding or three-dimensional interpretation of two-dimensional patterns. However, the reflections of the inner logic onto different contexts, such as the usage of different initial geometries, materials or performative concerns, were neglected. In this study, we offer a new schema to approach the performative aspects of muqarnas tectonics. This schema contains new sets of elements, properties and relations deriving partly from previous approaches and partly from the technique of folding. Thus, this study first reviews the previous approaches to analyse the geometric and constructional principles of muqarnas. Second, it explains the proposed scheme through a series of algorithmic form-finding experiments. In these experiments, we question whether ‘fold’, as one of the performative techniques of making three-dimensional forms, contributes to the analysis of muqarnas in both a conceptual and computational sense. We argue that encoding vault-like systems via geometric and algorithmic relations based on the logic of the ‘fold’ provides informative and intuitive feedback for form-finding, specifically in the earlier phases of design. While focusing on the performative potential of a specific fold operation, we introduced the concept of bifurcation to describe the generative characteristics of folding technique and the way of subdividing the form with respect to redistribution of the forces. Thus, in this decoding process, the bifurcated fold explains not only to demystify the formal logic of muqarnas but also to generate new forms without losing contextual conditions.

This study describes a teaching experience that aims to combine application-oriented approaches to computational design pedagogy. In the scope of this study, a hybrid framework is presented, which allows students to transition from manual to digital modes of design. In this transition, parametric thinking and design are selected as a mediator. This hybrid framework is being tested in an elective course of an international master's program in interior design. Columns are selected as the architectural elements for parametric design and experimental production. An "exhibition" is chosen to complete the design, production and assembly tasks. Preliminary exercises and three main assignments guide the process. The three assignments are organized as follows: parametric design and modeling of a single column (Assignment 1-Digital), parametric design and modeling of a family of columns (Assignment 2-Digital), the materialization of an individual column and a family of columns using 3D printers (Assignment 1 and 2-Digital+Manual), and designing the exhibition for displaying columns (Assignment 3-Digital+Manual), producing and setting up the exhibition (Assignment 3-Manual). The experiment with a hybrid framework shows that parametric design (digital) modes that students were unfamiliar with were integrated into the process of developing parametric columns, while exhibiting designed products using their existing skill sets (manual modes). The participants' inability to fully illustrate their technical knowledge of parametric design and digital fabrication in a single course, and the lack of tools and materials to facilitate 1:1 scale digital fabrication have emerged as the major limitations of the teaching experience.