Ye Wang
Join
Ye studied Computer Science and researched in volumetric shape language for 3D print at Computational Fabrication Group at MIT. After graduation, she continued her passion in creating design tools. She worked on various mesh-based design tools and online platforms at Autodesk. She then joined Onshape as Principal Engineer to develop the first full-cloud mechanical CAD. Last year, Ye co-founded Join with Andrew Zukowski and Drew Wolpert, with the goal to bridge design and construction.
Architects and contractors struggle to keep up with the growing and accelerating landscape of available building products and materials, even as ever higher performance standards and faster project delivery timelines make adopting new methods and materials a necessity. Knowing they exist is not enough -- to take advantage of the latest methods and materials properly the architect needs to understand their behavior, capabilities, and limitations in the context of a particular project. And key product choices often happen later in a project. Is it possible to quantify and compare cost and performance much earlier in the design process?
The core technology that Join makes this possible is a novel programmatic representation of building products and buildings based on research in programming languages in artificial intelligence domain. The major new capability Join offer a designer is the power of computational design to generate and rationalize without, in most cases, requiring that they learn a new programming language. Instead, they can draw and manipulate geometry in the direct interfaces they are accustomed to, and have data and design issues surfaced to them there. This rests on a domain-specific programming language capable of describing both a world of available products with constraints on their use and a model of a particular project (a concrete building design), together with an engine that operates on it to verify that the model meets the requirements and offer solutions where it does not.
In addition to streamlining the design development through construction process, these tools can improve overall outcomes: providing designers with more accurate cost and whole-building performance estimates earlier in the process can help them understand the implications of their decisions as they make them, avoid design errors, and reach better outcomes for owners and occupants. Pushing the industry toward realistic cost and performance estimates will help to reduce the enormous cost overruns (averaging 23% of project value in North America but often many times that for large public projects) that contribute to the risk and inefficiency of the industry, and enable us to embrace deep holistic as-built performance standards to move toward truly sustainable building.
Join is a startup located in the heart of San Francisco. The first product Retriever, programmatic representation for Revit building models, was released this January and well received by local architecture firms. Soon in summer, we will release the part for building products.
