|
Ralph Grabowski interviews Clifton Harness
|
|
| TestFit is unique among software programs in that urban planners can interactively design and change site plans, taking into account building densities, offsets, parking lots, road alignments, and more. (See figure 1.) The work is done parametrically, with a touch of generative design thrown in. "Let's get rid of placing tracing paper over satellite photos," ceo Clifton Harness told me that was his impetus.
Figure 1: Dynamically adjusting interior roadways to optimize land use
TestFit came about when Clifton Harness met Ryan Griege at the University of Texas, and found common ground in building Flash games; none were ever completed. Never mind. That formed the basis of a good working relationship that lasted to this day.
Mr Harness became an architectural intern working for a real estate firm, but didn't want a 9-5 job. His day job, however, gave him an idea for software that would make it easier to design sites for new home or office construction. To him, it looked like a math-based problem. Mr Griege wrote an initial program that accepted criteria for placing buildings optimally on sites, but one that only Mr Harness could use. Then Mr Griege got serious, and spent two years writing the core for TestFit, shipping a multifamily generator in 2017, and then a parking generator in 2018. (As a former transportation planning engineer, this part of the interview was exciting for me.)
It was a story of moving goalposts as the two fine-tuned a product when they weren't exactly sure what it would become. Mr Harness called real estate developers in Dallas to get feedback, and they named their company BuildingForge. The first release of the software was named "Residential Engine," shipping in the fall of 2017. They had two customers, and so Mr Harness cold-emailed 3,000 professionals across the country to grow sales. After eight months, they were finally able to pay themselves a salary, and then hire another programmer. They renamed the software TestFit.
The two decided they needed an injection of funding, and by coincidence VCs [venture capitalists] were wanting to meet them. The two were picky about who they would accept. Jesse Coors-Blankenship of Parkway Venture Capital seemed like a good fit, and so they accepted $2 million for hiring sales people, creating an international retail channel, finding a public relations firm (which resulted in this interview), and adding more programmers. They now have customers in six countries.
(Mr Coors-Blankenship is better know for selling his Frustum generative design software to PTC for $70 million, where he is now senior vp of technology, as well as part of Parkway Venture. worldcadaccess.com/blog/2018/11/exclusive-qa-with-frustum-ceo-jesse-coors-blankenship.html)
Here is an example of the thinking Mr Harness and his team put into the TestFit software. Its database health check looks for these attributes (see the dashboard lower portion of figure 2):
• Inline unit score: do units work well together by width?
• Angular score: do units work well when turning corners, especially irregular corners?
• Target area score: do units achieve a wide variety of unit averages?
Figure 2: Dashboard reporting results in realtime
As you adjust units, the dashboard reflects the changes, whether positive or negative. After three years of development, TestFit offers these functions:
- Unit editor for defining building units
- Shadow casting
- Land planning
- Varied retail space
- Parking garages
- Zoning
It supports up to 16 categories of building use, along with configurable hallways, balconies and bays, stairwells, and so on. See figure 3.
Figure 3: Optimizing varied uses of a building
Generative design is used by recent marketing descriptions for TestFit, but to me it looks like the software primarily uses parametrics. Mr Harness said he likes to call his version "practical generative design" in that the software places things like staircases as efficiently as possible when generating building masses. (It's about 10% of the code base.) You specify a minimum number of staircases, constraints with parameters, and TestFit works within it.
He doesn't like the voluminous output from typical generative design solutions for BIM. His goal is for his software to generate, say, two really good buildings, instead of 15,000. "Building design is tough for generative design, as people have to live inside of them," he said. Generative design is meant to optimize a design, but there is a lot of core technology like the staircases before the building can be optimized.
A building is a compromise between many parameters, with which TestFit works with faster than does tracing paper. You cannot optimize buildings for just one parameter, such as cost (no one would want to live there) or courtyard space (too expensive to build). And it needs to keep the initial design cost down, because only one out of ten building designs eventually get built.
The data is stored in its own format with some meta data attached to each entity. The results can be output in PDF, DXF, SKP (SketchUp), glTF (graphics language transmission format for 3D scenes and models using JSON), RSD (TestFit's native format), or CSV (comma-separated value) formats.
"Are architectural CAD vendors interested in TestFit," I wondered. From the response, I gather they are, but it is too soon to talk about it.
"Who is your typical customer?" I asked. Mid-size real estate development firms, architecture firms of any size that want to jump into new technology, and more recently general contractors who see it useful for quantity take-offs.
You can download a 30-day demo from https://www.buildingforge.com/ after registering.
https://blog.testfit.io/
|
|
Sponsor: Okino Graphics
== 3D Conversion of Ultra-Massive 3D Models via DWF-3D
& Okino's PolyTrans|CAD ==One of the most refined aspects of Okino's PolyTrans|CAD software is in transforming ultra-massive MCAD models of oil and gas rigs, LNG processing plants, 3D factories, and other unwieldy datasets into Cinema-4D, 3ds Max, Maya, and Unity (among others). What often takes days using blindly incorrect methods takes minutes or an hour with Okino's well-defined optimization and compression methods using its DWF-3D conversion system. Popular CAD data sources include SolidWorks, ProE/Creo, Inventor, AutoCAD, Revit, Navisworks, DGN, IGES, STEP, Parasolid, and JT. DCC data sources are Cinema-4D, 3ds Max, Maya, FBX/Collada, and many more.Perfected over three decades, we know 3D data translation intimately, providing you with highly personalized solutions, education, and communication. Contact CTO Robert Lansdale at [email protected] http://www.okino.com
|
|
Re: Readers Respond to When Generative Design Backfires
In response to the person that said "comparing a spoked bicycle with a car wheel makes no sense." Perhaps a more apples-to-apples comparison then would be with an older car with wire wheels.
-Joe Tilman
The editor replies: You know, I was thinking about spoked wheels on cars, but then I forgot to mentioned them. I am glad you reminded me. I found that spoke wheels are still available today.
- - -
It always amazes me what kind of real-world issues crop up that a designer can't foresee. A friend of mine was griping about aluminum wheels and said that his company now swapped out the aluminum wheels that come on new pickups for steel one, because the aluminum ones wear out.
He finally clarified that, in constant operation on dirt roads, the abrasive dust and sand gets between the tire sidewall and the rim for just that tiny little band where the rubber flexes away from the rim, and gradually wears the rim of the wheel away until either the tire comes off, or, more likely, the rim breaks when they try to install new tires.
I've been around farm machinery, construction machinery, dirt bikes, and pickups most of my life and never heard of that, but it's only very recently that pickups in that kind of service (oilfield maintenance) have ever had aluminum wheels.
- Jess Davis
The editor replies: All that aluminum is a desperate attempt to bring fuel consumption levels down to what politicians fanaticize should be feasible.
Mr Davis responds: Ever since I got a couple of nice digital scales for a project, I've been weighing every wheel that I get in my hands. It's astonishing to me how terrible aftermarket aluminum wheels are, the few I've had. My notes are scattered all over a couple of different devices, but I remember switching some rims around on my daughter's Nissan, and the aftermarket alloy rims I had laying around pretty much weighed almost exactly what the factory Nissan steel ones did -- 20 pounds, I think. By comparison, the stock 14" rims from my old Miata were 11.5 pounds, way lighter than most aftermarket "racing" wheels.
3/4 ton Dodge stock alloys were extremely light compared to stock Dodge steels.
Racing makes a person obsess about weight, and growing up with actual, real race motorcycle bikes apparently made me kind of snobbish. Now that I'm too old for the real fun, and have to content myself with street-legal toys, I'm often appalled at the weight of components.
I remember the first bike I had that got heavier while I owned it. There was a kind of tipping point back in the late 80s when Honda designed past the fatigue limit for a lot of stuff, and when you installed aftermarket parts, they were always heavier than OEM. That was weird. Finding fatigue cracks on things like rims was kind of expected, but the cracks on stuff like fork clamps and handlebars was alarming.
|
|
Thank you to readers who donate towards the operation of upFront.eZine:
To support upFront.eZine through PayPal, then the suggested amounts are these:
Should Paypal.me not operate in your country, then please use www.paypal.com and the account of [email protected].
Or mail a cheque (US$ or CDN$ only, please) to upFront.eZine Publishing, Ltd., 34486 Donlyn Avenue, Abbotsford BC, V2S 4W7, Canada. |
|
|
|
Comments