Machine Learning Now! How Rhino Inside plus Hops brings AI to Revit
Matthew Breau | Computational Design Lead
Handel Architects | https://handelarchitects.com/
Machine Learning (ML) has matured into an ecosystem of robust open-source frameworks. However, Machine Learning’s industry-standard tools are siloed off from AEC design tools, making it difficult to conduct R&D of ML in AEC. As AEC hackers and tinkerers, how can we take development of AEC-ML tools into our own hands? What would this even look like?
In this presentation, we share the findings of Handel Architect’s R&D, using existing Rhino Inside and Rhino Hops technologies to connect heterogeneous tools across the AEC and ML fields. This support's Handel Architects goal of creating cutting edge tools to assist designers, by leveraging existing top-shelf open source frameworks and tools.
The presentation will review the basic value proposition of Machine Learning, and introduce Handel Architects’ research, which focuses on pattern matching for “autocomplete” style design suggestions, task automation, and BIM model quality assurance. We will also present an ML case-study which demonstrates how Handel Architects is combining off-the-shelf tools and frameworks to: extract BIM data into ML-friendly datasets, train and provision an ML model, query that model, and use the results in Revit. Finally, the presentation will lay out a vision for further development of ML-based AEC design tools, outlining how deploying simple tools now can position firms to develop more advanced tools in the future, by building high-quality datasets.
Learn more about AECtech events at: https://www.aectech.us/
© 2021 CORE studio at Thornton Tomasetti