TestFit on Generative Design

Schemes Tool Testfit to compare site plans

A good friend, Anthony Hauck, wrote an exhaustive explanation of what generative design is. I will stick with his definition of generative design (stand on the shoulders of giants, right?).

Generative design is the automated algorithmic combination of goals and constraints to reveal solutions.

We are a practical startup. We build tools users can use right now and can implement into their workflow today. Over the last 2+ years of working on TestFit, we have been incrementally working towards building-level generative design. Along the way we took some detours for our customers and built things like manual mode, the unit editor, and multiple sites. But the long-term goal has been building-level generative design.

Introducing Schemes

With TestFit, our thinking was to create one quality scheme, and to do it in a few milliseconds. This way users could rapidly prototype different schemes. One thing that our users requested was the ability to compare different schemes easily. With our schemes tool, users can now do that!

Schemes Tool Testfit to compare site plans

Filtering Schemes

Invariably, when presented with more than a few options, it becomes hard to sort out what is good and what is bad–and we have a hard limitation: the human mind can really only compare four things at once. This is why data-driven design tools typically have a parallel coordinate graph to filter out solutions that are not ideal, and to present a narrow range of options. We have built our own parallel coordinate graph, comparing design, zoning, and financial information.

parallel coordinate graph, comparing design, zoning, and financial information

Automatically Generating Schemes

The blueprint for this started with the first version of TestFit. With the 1.0 update, we launched presets. Presets enable users to create their own lists of parametric values, and to simply click between them (think 5 over 2 or type III wrap). TestFit also has a feature called options, which are simply different building mass layouts for the site. The cross-product of “presets” and “options” revealed that there were already several schemes available for users. Here is what automatically generated schemes looks like:

Automatically generated schemes in TestFit

We think, for starters, that 64 schemes should be more than enough. We currently exist in a world where fewer than three schemes are ever test-fitted per site. With a 20-fold increase, we should be able to find more opportunities to solve real estate deals, or at least the data to kill them.

Generative Design for Parking Garages

Generative parking

Why build generative design for parking garages? What do we build most in this world? Parking… What do we hate most in this world? Parking… How many stalls have you counted in the last 30 days? Too many.

Ugh

Parking can be 20% to 70% (by area) of any project that is built. It’s almost shameful when half of your project is parking, but here we are as a culture–there are 8 parking stalls for every 1 car in this country. Parking is a basic need for any building to be successful–so we design it, build it, and tolerate it. The more efficient the parking solution, the better. This means attempting rectangular garages, parking on ramps, and a stall average south of 400 square feet per stall.

Basic Parking Garage Parameters

  1. Stall Depth – We default to 18′ (we are based in Texas, after all)
  2. Stall Width – We default to 9′ (see reason above)
  3. Drive Aisle Width – We default to 24′ (trucks are large)
  4. Max Ramp Slope – We default to 6%
    • Building code in all of its glory mandates a maximum of 6.66% slope (15 feet of run for every foot of fall). This limits your ability to squeeze stalls, but for a good reason (the guys writing the code were thinking about people in wheelchairs?).

Basic Parking Garage Objectives

  1. Draw the stalls given the user’s inputs
  2. 95% of the garage drawn (in milliseconds)
  3. What bay configuration yields the most parking stalls
  4. What ramp location kills the fewest stalls
  5. Fill any irregular shape with parking

The TestFit Generative Design for Parking Solution

These features are now live within the main TestFit app. For future features to do with parking? We are considering speed ramps, spaces within the garage, entry and exit locations, angled parking, and directional routing. Our intent is to give the industry a tool to solve 80% of garages, and do it in milliseconds.

TestFit: Startup Year One In Review

Startup year in review

TestFit was founded to design buildings in milliseconds–to help architects and developers solve real estate deals with ease. This idea, and our implementation of it, has enabled us to survive our first year as a startup. As we approach October 1st, the one year anniversary of releasing TestFit, we are taking a moment to look back over the past year on all that we have accomplished with our Startup Year One in Review.

October 2017

  • Launched Residential Engine – The first out-of-the-box generative design tool
  • Made the first sale of Residential Engine

November 2017

  • Released Residential Engine 1.011 – “Spaces and 3D”
  • First outside of Texas sale (Atlanta, Georgia)

December 2017

  • Released Residential Engine 1.012 – “Tabulation”
  • First out of state trip (Atlanta, Georgia)
  • First annual contract, company-wide sale

January 2018

  • Released Residential Engine 1.013 – “Input”
  • Formalized macro-bim as a possible value proposition
  • Presented Residential Engine at Society for Construction Solutions, Boston
  • First Sale in Florida

February 2018

  • Released Residential Engine 1.014 – “Manual Mode
  • Presented Residential Engine at Society for Construction Solutions, San Francisco
  • First Sales in California and Massachusetts

March 2018

  • Released Residential Engine 1.015 – “Wrap-Podium”
  • First Sales in Colorado and Wisconsin

April 2018

  • Re-branded to TestFit
  • Released Test Fit 1.0
  • First Sales in Missouri, Oklahoma, Pennsylvania, Minnesota and Virginia
  • Formalized Deal Information Modeling
  • First International Sale to the United Kingdom

May 2018

  • Released TestFit Version 1.001 – “Multiple Sites”
  • Implemented Deal Information Modeling
  • First sales in Indiana and Washington

June 2018

  • Released TestFit Version 1.002 – “Lifts”
  • Drew up some real estate infographics
  • First conference booth at AEC Next
  • Won AEC Hackathon for “Scan Point”
  • First sales in Arizona and France

July 2018

  • Released TestFit Version 1.003 – “Unit Editor”
  • Presented TestFit at USC BIM BOP – Our first academic engagement!
  • First Sales in Maryland and Iowa

August 2018

  • Released TestFit Version 1.004 – “Dynamo”
  • Released Submit-A-Site – A SaaS solution for the test fit process
  • First company to publish an algorithm on Hypar
  • Posted our first job listing!
  • Submitted Generative Noise Abatement to NMHC
  • First Sales in North Carolina

September 2018

  • Working on a secret project to be released sometime in October
  • First BIM model built from TestFit geometry
  • First implementation of automatic free trials for TestFit
  • Accepted to Y Combinator Startup School