Workshops
DepthSpace3D — a digital tool for 3D Space Syntax
Objectives
The workshop will give a first approach to the practical use of DepthSpace3D.
Participants will be able to develop spatial analysis of projects in the areas of Architecture, Urban Planning and Landscape Architecture, optimizing solutions in security (CCTV placement, agent placement, children surveillance), economy (value assignment to space, product placement, publicity), traffic (accessibility, clutter, jams prediction), aesthetics (monumentality, space diversity, remarkable places), urbanism (skyline protection, centralities, equipment distribution), and social studies (social segregation, control, functional uses of space).
This know-how has a professional demand not only in the architectural office, but also in real-estate, large buildings’ management or public administration.
Expected Outcomes
The course is short and focused on DS3D, and does not contain a theoretical approach of Space Syntax methodologies. This situation implies that students must have previous knowledge of Space Syntax and its usual computer applications.
By the end, participants will have an idea of DepthSpace3D's capabilities in addressing concrete problems and will be able to begin operating the software application at an elementary level, which they can develop further in the future.
Reality Overlap Lab: From Embodied Micro-Place Annotation to Formal Spatial Analysis
Overview
Reality Overlap Lab is a half-day, hands-on workshop and a structured component of Oğuz Emre Bal’s doctoral research in Architectural Design at Istanbul Technical University, supervised by Prof. Dr. Pelin Dursun Çebi. The workshop will be conducted in the ITU Taşkışla Central Courtyard and will apply the Reality Overlap method to investigate how embodied, first-person accounts of architectural space can be captured, structured, compared, and formally analysed.
Participants will move through the courtyard and identify one or two personally meaningful micro-places. Using a Meta Quest 3 headset, they will place persistent spatial anchors and record short situated narratives. Each record will connect the participant’s testimony to a calibrated three-dimensional coordinate, viewpoint, timestamp, movement trace, media record, and visibility status.
The captured traces will then be returned to a digital twin and examined through four analytical constructs: micro-spatial meaning density, semantic diversity, interpretive overlap, and trace-mediated social reorientation. Working in small groups, participants will inspect AI-assisted thematic interpretations, formulate transparent operational rules, compare interpretive distributions with movement data, and identify convergence, divergence, ambiguity, and methodological edge cases.
The spatial anchors, first-person narratives, movement traces, thematic interpretations, and participant reflections generated during the workshop will provide methodologically valuable empirical material for evaluating and refining the doctoral study’s methodological and analytical framework. Only data covered by explicit informed consent and the relevant institutional ethical requirements will be retained, analysed, or used in subsequent doctoral research and publications.
Objectives and Learning Outputs
By the end of the workshop, participants will be able to: explain the operational definitions of a micro-place and a persistent spatial anchor; describe how spatial meaning can be captured in situ through coordinates, viewpoint, timestamp, movement trace, voice narrative, image, and visibility status; inspect and critically review human-validated, AI-assisted thematic interpretations of free-form spatial narratives; construct transparent operational rules for micro-spatial meaning density, semantic diversity, interpretive overlap, and trace-mediated social reorientation; compare interpretive distributions with movement intensity and discuss why spatial significance may not coincide with circulation frequency; identify convergence, divergence, ambiguity, and edge cases within multi-user spatial data; critically discuss privacy, authorship, AI interpretation, researcher positionality, and the limits of formalising lived spatial experience; and consider how the workflow may be transferred to educational, residential, healthcare, heritage, and civic environments.
Expected Outcomes and Research Value
The workshop will produce participant learning outcomes and a small but methodologically valuable spatial dataset as a structured component of the doctoral research. Expected outputs include:
- an anonymised or pseudonymised set of spatially registered first-person micro-place narratives;
- an annotated digital-twin view combining anchor locations, movement traces, timestamps, media, and thematic interpretations;
- participant-generated operational definitions for meaning density, semantic diversity, interpretive overlap, and trace-mediated social reorientation;
- a preliminary comparison between movement intensity and interpretive activity, used as a methodological exercise rather than as a statistically generalisable result;
- examples of cross-user convergence, divergence, and semantic ambiguity at the same or adjacent micro-places;
- a documented list of methodological and ethical edge cases, including spatial-threshold selection, theme ambiguity, AI misclassification, privacy, researcher influence, and the distinction between representation and measurement; and
- a transferable workflow that participants may adapt to other architectural research contexts.
The workshop will not be treated as a source of statistically generalisable or confirmatory findings. Its research value lies in producing a structured empirical setting through which the Reality Overlap protocol, its analytical constructs, and its practical limitations can be examined and refined as part of the doctoral study.
From Prompt to Fabrication: Integrating Generative AI, Real-Time Media, and Computational Fabrication in Architectural Design
Overview
This workshop explores an integrated architectural design workflow that brings together three node-based programming tools: ComfyUI, TouchDesigner, and Rhino-Grasshopper. Participants will begin by generating architectural imagery from text prompts in ComfyUI, transform these images into computational three-dimensional forms in TouchDesigner, and subsequently adapt the resulting geometries into fabrication-oriented simulation models in Rhino-Grasshopper.
Through the integration of these platforms, the workshop investigates how design information can move across different computational environments, linking language, image, geometry, and fabrication within a continuous architectural design process. The workshop emphasizes experimentation, iterative development, and the translation of information between systems. Participants will gain practical experience with generative AI, real-time computational design, and computational fabrication while exploring contemporary approaches to digital architectural production.
Objectives
By the end of the workshop, participants will:
- Generate architectural imagery using AI-assisted text-to-image workflows.
- Transform visual information into computational three-dimensional forms through real-time generative processes.
- Create fabrication-oriented simulation models using parametric design tools.
- Develop a foundational understanding of node-based programming logic and visual scripting workflows.
- Gain experience integrating generative AI, real-time media, and computational fabrication within a connected architectural design workflow.
- Understand how design information can be translated across different computational environments.
Expected Outcomes
By the end of the workshop, each participant will produce:
- A series of AI-generated architectural images using ComfyUI.
- A computationally generated 3D form and real-time renders derived from those images in TouchDesigner.
- A robotic fabrication simulation prepared in Grasshopper.
- A documented workflow demonstrating the transformation of design information across the three platforms.
Modernity of Capriccio. Ruin as a Contemporary Project
Overview
Participants must conduct, from a photographic image, an analysis by sections of the different layers that make up the urban landscape. These sections will then have to be assembled and brought back to a collage, that is, a small-scale model (35x35 cm) in which to recompose the various layers identified in the urban landscape. Each participant will be given a photograph of a city and a contemporary building, which they must transform into a ruin with the aim of understanding its formal, structural, and symbolic values.
The workshop fits within the Formal Methods in Architectural Education line of research, as it combines the technique of digital and analog collage with that of automatic drawing reproduction. The work may be conducted freehand or with the aid of digital instrumentation and will lead to the creation of a small capriccio. The ultimate aim of the workshop is to create a synthesis between the urban landscape and the form of the building, enabling students to understand how the contemporary city, too, can be the result of transformations, layers, and processuality.
Objectives
A small "collage" inspired by Giovanbattista Piranesi, Giovanni Paolo Panini, Hubert Robert and Luigi Rossini works, in which to place all the sections identified by the layered reading of urban landscapes.
Robotic Casting with Flexible Fabric Formwork
Overview
This workshop introduces a robotic concrete casting approach grounded in tactile fabric exploration and computational making, translating hands-on observations of material behavior into parametric robotic fabrication strategies. It invites participants to explore how robotic twisting, tensioning, and timed intervention can become geometric procedures within a flexible fabric formwork process.
The workflow focuses on twisted fabric formwork as a generative medium. Within the stretched and rotated fabric, concrete records the interplay of material weight, elastic resistance, timing, and robotic motion. The resulting modules carry traces of this controlled forming process through curvature, cross-sectional variation, local thinning, surface deformation, and subtle irregularities.
Using Grasshopper and KUKA|prc, participants will translate a selected forming action, first explored through hands-on fabric experiments, into motion parameters such as rotation angle, pulling direction, tool orientation, tension, and intervention timing. Each team will develop a unique module that can be discussed as part of a larger assembly system.
Objectives:
By the end of the workshop, participants will be able to:
- understand robotic casting with flexible fabric formwork as a computational making process;
- relate fabric behavior, concrete setting, and robotic motion to geometric outcomes;
- translate tactile observations of twisting, elasticity, tension, deformation, and timing into parametric motion strategies;
- examine how material-forming actions generate variations in curvature, cross-section, and surface character;
- produce an individual concrete module through controlled robotic casting;
- discuss how this module may inform larger aggregated architectural scenarios.
Expected Outcome
The workshop will result in a set of individual concrete modules produced through different robotic motion scenarios. These outputs will demonstrate how variations in tension, rotation, timing, and tool movement affect the geometry, cross-section, and surface character of cast concrete elements.
The final discussion will invite participants, workshop leaders, and conference attendees to reflect on the potentials and limitations of flexible fabric formwork, material-driven robotic control, modular assembly, and sustainability considerations in concrete casting workflows.
Form-Finding and Bio-Integrated Simulation of Structures through Grasshopper, Kangaroo, Karamba3D and Generative AI Imaging Technologies
Objectives:
Part 1 — Funicular Form-Finding and FEA (Finite Element Analysis) Simulation in Kangaroo and Karamba3D: Teach participants to generate compression-only vaulted geometries in Grasshopper using Kangaroo 2, applying stereo-funicular equilibrium principles inspired by Gaudí. Convert the resulting geometries into structural FEA models in Karamba3D to analyze axial forces, bending moments, shell stresses, and deformations under self-weight and permanent loads from roof soil/biomass.
Part 2 — Predictive Simulation through Generative AI Image Engines: Introduce generative AI image-synthesis workflows (such as latent-diffusion-based AI engines, e.g., Stable Diffusion or Midjourney) using depth maps and geometries generated in Grasshopper to visually simulate the final "Built & Born" state, forecasting long-term biological colonization (mosses, lichens, plant roots) as well as the degradation and erosion resulting from weathering action on the material over time horizons of 1 to 10 years.
Part 3 — Physical Prototyping and Mesh Unrolling: Provide hands-on experience in rationalizing double-curvature meshes into flat, printable two-dimensional models (rigid origami/paneling logics) for assembling scale models in paper or cardstock, validated by Karamba3D structural analysis.
Expected Outcome
Participants will master an end-to-end computational workflow linking physical form-finding, structural validation through Finite Element Analysis (FEA) in Karamba3D, and generative AI visual simulation. Key deliverables include a fully functional Grasshopper/Kangaroo/Karamba3D script for structural analysis under earth/biomass loads, an AI diffusion pipeline that forecasts long-term biological maturation, and a physical cardstock prototype at 1:200/1:50 scale (depending on the project).