TY - GEN A1 - Yazdi, Hadi A1 - Banach, Marzena A1 - Moazen, Sajad A1 - Nadolny, Adam A1 - Starke, Rolf A1 - Bazazzadeh, Hassan T1 - Central Courtyard Feature Extraction in Remote Sensing Aerial Images Using Deep Learning: A Case-Study of Iran T2 - Remote Sensing N2 - Central courtyards are primary components of vernacular architecture in Iran. The directions, dimensions, ratios, and other characteristics of central courtyards are critical for studying historical passive cooling and heating solutions. Several studies on central courtyards have compared their features in different cities and climatic zones in Iran. In this study, deep learning methods for object detection and image segmentation are applied to aerial images, to extract the features of central courtyards. The case study explores aerial images of nine historical cities in Bsk, Bsh, Bwk, and Bwh Köppen climate zones. Furthermore, these features were gathered in an extensive dataset, with 26,437 samples and 76 geometric and climactic features. Additionally, the data analysis methods reveal significant correlations between various features, such as the length and width of courtyards. In all cities, the correlation coefficient between these two characteristics is approximately +0.88. Numerous mathematical equations are generated for each city and climate zone by fitting the linear regression model to these data in different cities and climate zones. These equations can be used as proposed design models to assist designers and researchers in predicting and locating the best courtyard houses in Iran’s historical regions. KW - central courtyard KW - deep learning KW - features extraction KW - data analysis KW - iran Y1 - 2021 UR - https://www.mdpi.com/2072-4292/13/23/4843/htm# SN - 2072-4292 VL - 13 IS - 23 ER - TY - GEN A1 - Starke, Rolf A1 - Vukorep, Ilija A1 - Frommelt, Konrad A1 - Melcher, Alexander A1 - Hinze, Thomas ED - Frier Hvejsel, Marie ED - Cruz, Paulo J.S. T1 - Towards artificial ossification for bone-inspired technical structures T2 - Structures and Architecture : A Viable Urban Perspective? N2 - Since its first description in 1892, the adaptation of internal bone structure to changing loading conditions over time, known as Wolff 's Law, has inspired a wide range of research and imitation. This investigation presents a new bone-inspired algorithm, intended for the structural design of technical structures and capable of optimizing the shape and size of three-dimensional lattice structures. Unlike conventional structural optimization methods, it uses interacting artificial agents that closely follow the cellular behaviour of the biological blueprint. Agents iteratively move, alter cross-sections, and reposition axes in the latticework. The efficacy of the algorithm is tested and evaluated in two case studies. This agent-based approach lays the theoretical foundation for an implementation of adaptive structural building components and provides a tool for further research into the spatial aspects of natural ossification. KW - artificial ossification KW - architecture KW - structure Y1 - 2022 SN - 978-0-367-90281-0 U6 - https://doi.org/10.1201/9781003023555-145 SP - 1211 EP - 1218 CY - Aalborg ER - TY - GEN A1 - Kotov, Anatolii A1 - Starke, Rolf A1 - Vukorep, Ilija T1 - Spatial Agent-based Architecture Design Simulation Systems T2 - Co-creating the Future: Inclusion in and through Design - Proceedings of the 40th Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2022) KW - Agent Simulation KW - Reinforcement Learning KW - AI Aided Architecture Y1 - 2022 UR - http://papers.cumincad.org/cgi-bin/works/paper/ecaade2022_176 PB - Ghent ER - TY - GEN A1 - Vukorep, Ilija A1 - Fiebig, Jan A1 - Starke, Rolf A1 - Eisenloffel, Karen ED - Kontovourkis, Odysseas ED - Phokas, Marios ED - Wurzer, Gabriel T1 - Applied artificial ossification for adaptive structural systems T2 - Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024) N2 - This study explores the "Artificial Ossification" algorithm's application in real-world structures, inspired by human bone formation. It uses agents mimicking bone-building and degrading cells to iteratively optimize structures for equilibrium through the Finite Element Method. The research proposes a 3D printing pen method for material addition or removal, mirroring natural bone adaptability and sustainability. Initial tests on 3D-printed models showed promising results, leading to more rigorous comparisons between conventional and algorithm-optimized structures. Findings confirm the algorithm's practicality for adaptive, optimized structural design, with potential applications in architecture, engineering, and beyond. The study also highlights the method's sustainability, repairability, and scalability, suggesting its relevance for future research in adaptive materials and design methods. KW - Artificial Ossification KW - Adaptive Structural Systems KW - Bionics KW - Shape Optimisation KW - Bone Inspired Structure Y1 - 2024 UR - http://ecaade.org/current/wp-content/uploads/2024/09/eCAADe2024_Volume1-r.pdf SN - 9789491207372 SN - 2684-1843 SP - 75 EP - 84 CY - Nicosia, Cyprus ER - TY - GEN A1 - Vukorep, Ilija A1 - Starke, Rolf A1 - Khajehee, Arastoo A1 - Rogeau, Nicolas A1 - Ikeda, Yasushi T1 - GPU-accelerated collision-free path planning for multi-axis robots in construction automation T2 - Proceedings of the 42th International Symposium on Automation and Robotics in Construction (ISARC) N2 - The Architecture, Engineering, and Construction (AEC) sector faces increasing pressure for higher production rates amidst a growing shortage of skilled labor, driving the demand for advanced robotic applications to enhance precision, efficiency, and adaptability in complex environments. This paper introduces a software setup designed to ensure collision-free movements for multi-axis robots in AEC scenarios. Our approach leverages the NVIDIA cuRobo framework's robust capabilities, seamlessly integrated with Grasshopper for Rhino 3D software (GH), a tool widely recognized for its versatility in parametric design. The integration of these technologies allows for the efficient online generation of optimal path movements, avoiding collisions even in highly intricate settings and changing environments. This is achieved in a remarkably short timeframe, enhancing productivity and reducing downtime. NVIDIAs framework's GPU-driven architecture paired with our GH parametric and controlling setup is a significant advancement, validated through a case study involving a complex, tree-like structure constructed from timber sticks. Using a six-axis robotic arm, the study demonstrates the system's capability to navigate and manipulate within congested spaces efficiently. With this enhanced automation workflow, new possibilities emerge for robotic applications, from industrial automation to sophisticated construction projects. Our GH software also allows visualization and exchange with URDF-models and better planning of collision logic, which was previously only possible with ROS and Nvidia Isaac technology. KW - Collision-free path planning KW - Multi-axis robots KW - AI robotic automation KW - Parametric design Y1 - 2025 SN - 978-0-6458322-2-8 U6 - https://doi.org/10.22260/ISARC2025/0056 SN - 2413-5844 SP - 421 EP - 427 PB - International Association for Automation and Robotics in Construction CY - Montreal, Canada ER - TY - GEN A1 - Fritzsche, Lukas A1 - Starke, Rolf A1 - Felbrich, Benjamin A1 - Vukorep, Ilija T1 - Evaluation framework for indoor localization systems in the AEC environment T2 - ISARC 2025 : proceedings of the 42th International Symposium on Automation and Robotics in Construction N2 - Accurate spatial localization is critical in the construction industry, indoor robotics, and other applications involving actor movement within buildings. This paper introduces a combined software and hardware framework designed to evaluate indoor localization systems leveraging optical tracking integrated with a ROS communication system. It supports recording location data, aligning trajectories and evaluating the recorded samples against the ground truth allowing for real world comparison within a custom setup. While it is designed to evaluate any localization system, a presented case study compares two established SLAM algorithms regarding their suitability for an AEC application. KW - Localization evaluation KW - Indoor localization KW - Mobile robotic KW - SLAM KW - ROS Y1 - 2025 SN - 978-0-6458322-2-8 U6 - https://doi.org/10.22260/ISARC2025/0067 SN - 2413-5844 SP - 508 EP - 515 PB - International Association on Automation and Robotics in Construction CY - Montreal, Canada ER -