Generative Design at BILT North America 2019
A defensive pessimist at heart, I was pleasantly surprised and greatly encouraged by the response to the call for generative design abstracts for BILT North America this year. Far from having to shoehorn some computational classes into the new category, for a single stream the committee enjoyed an embarrassment of riches, to the point where we have more than a full schedule throughout the conference. If your generative design class wasn’t chosen this time around, the committee would encourage you to update it and submit it again for future years and venues. We expect interest in this area to grow year after year.
In the final schedule for this summer we have a great mix of generative design practices within professional firms, presentations of commercial and pre-commercial products, roundtable discussions, and a generous quantity of hands-on labs to help make this newly practical way of working more accessible.
We start the conference with two successive labs focused on machine learning, a predictive computational approach I expect will be routinely applied to enhance project understanding, decision quality, and generative design workflows within just a few years. Radu Gidei brings us Practical Applications of Machine Learning in Dynamo, demonstrating how to use visual programming with data science techniques for predictive analytics, followed by Adam Sheather with Machine Learning: Getting Started with AEC Data, bringing the audience from the ground up to automating estimation of clash importance. But hurry these classes are limited with only a few BYOD spots left!
Alongside the second half of Adam’s extended lab, Lilli Smith will present Generative Design with Project Refinery, which rolls directly into the end of day one with a lab from Matt Jezyk on How to Train Your Model with Project Refinery, helping attendees first get familiar and then get started with hands-on generative optimization on the desktop, leveraging Dynamo to converge on performative solutions.
On Day Two we kick off with Generative Design Using the Hypar Platform from my business partner Ian Keough, a lab offering hands-on experience with optioneering and generative cloud computing to attack a variety of building problems.
Directly following Mark Thorley presents Early Stage Feasibility Assessments Through Generative Design while Chris Tisdel conducts a roundtable on Integrating Generative Design Innovations into Practice. Afterward Clifton Harness will present TestFit: Generative Design and Real Estate Development, focusing on modeled feasibility studies with multi-variable optimization and interaction.
The day winds up with another lab, Mars Capsule Living Quarters Design with the Aid of Generative Design from Dat Lien, bringing the audience from optimizing solutions for residential space travel on Elon Musk’s Big Falcon Rocket to immediate uses in building design decisions here on Earth.
Day three opens with Zach Kron’s Farming Forms and Managing Minions with Generative Design, where he’ll explore common underlying approaches of the generative design discipline while likening computational processes to a junior designer working under guidance. From there we move on to Cabin Fever: Generative Design for a Vernacular Classic from Kyle Martin, taking the audience on his journey from classic gestural design workflows to incorporating generative design into practice using a small building example to illustrate techniques.
Next Achintya Bhat presents Architectural Model to Schematic Structural Layout using Generative Design, demonstrating best practices in evolving structural systems from architectural data and captured engineering expertise. We finish the conference and the generative design stream with Nathaniel MacDonald’s Generative MEP in HYPAR where he’ll employ architectural requirements to deliver generative MEP systems on the open Hypar platform while Juhun Lee facilitates a roundtable on Reinforcement Learning Models for Engineering Design Problems, where he’ll demonstrate techniques for training predictive models with a minimum of input data applied to real structural design problems.
The BILT North America committee couldn’t be more pleased with the strength and diversity of the submissions for this inaugural Generative Design stream of classes. With such a compelling roster of presentations, discussions, and labs we’re sure you can leave the conference with a solid grounding in generative design that you can bring to your projects, capturing and extending your firm’s expertise to augment the speed and quality of the decisions you and your firm make every day. We’re looking forward to seeing you in Seattle, where our presenters will be pushing the boundaries of modern design and engineering across new frontiers!
Check out the full schedule for this year’s Digital Built Week North America here