설명
Integrating building performance analysis results in the early stage of conceptual design supports effective design exploration and decision making. Such integration requires domain expertise and can be technically challenging, time consuming, and computationally expensive. Machine learning demonstrates great potential for building seamless workflows that provide faster analysis feedback. However, one of the challenges in the architecture domain is the lack of diverse data sets to train effective prediction models. This talk will provide a brief overview of the potential of machine learning in building analysis-prediction models and demonstrate how we can capitalize on generative design to compensate for the lack of data. We will present a workflow where generative design features for Revit software, Dynamo software, and Insight software are used to generate a synthetic data set for training a machine-learning model for energy-analysis prediction.
주요 학습
- Learn how to design workflows with Generative Design in Revit and Dynamo for building synthetic data sets to be used in training machine-learning models.
- Discover the diversity of mass model geometry required to represent a comprehensive set of possible building types.
- Learn how to represent your data to be used in training machine-learning models.
- Discover potential uses of machine-learning models toward achieving faster analysis in early conceptual design stages.
발표자
Varvara ToulkeridouVarvara Toulkeridou is a Sr. Research Engineer for Autodesk’s AEC Generative Design group, currently working on Generative Design for Revit. Her role at Autodesk has focused on enhancing generative design workflows with software tools that facilitate the generation and evaluation of design options. In parallel, she is pursuing a Ph.D. in Computational Design at Carnegie Mellon University. Her doctoral research focuses on using machine learning to augment parametric modeling tools to support design exploration. She holds an MSc in Design and Computation from Massachusetts Institute of Technology and a Diploma in Architecture from the Aristotle University of Thessaloniki. She has taught and assisted courses at CMU and MIT on digital media, parametric modeling and scripting, descriptive geometry, and digital fabrication. She has presented her work and taught workshops at the eCAADe and SIGraDi conferences.- MFMichael FloydMichael Floyd is an environmental technologist and design director, accelerating solutions that advance environmental & social performance in the AEC sector—for a sustainable hardscape. Current focus on high performance buildings, Lean and industrialized construction, and resilient cities. Recently led integration of EC3 with BIM 360 for carbon-smart building material choices. His life prior to Autodesk included design direction and consulting in both product development and the built environment. His product-oriented work has addressed toxicity, design for environment, climate, and supply chain adaptation. He has a special interest in energy productivity in the built environment, and has undertaken related work at scales ranging from whole cities to commercial kitchens. He holds an MSc in Environmental Technology from Imperial College London, and a BA from Princeton University.
Video Player is loading.
No transcript