InstantCAD AI research generates industrial design on demand
MIT researchers have launched a tool that automatically generates products and analyzes them in detail on your behalf.
Take these two work lights. They each have three heads, curved and placed in very different ways. So who has the best stability? This is a trick question. They are also stable, and this was discovered by an algorithm that designed them both.
Researchers at MIT, in collaboration with Columbia University, have unveiled a new tool for designers who work with computer-assisted drawing software. Building on previous work from the past year, their technique can optimize the design of any object, such as a lamp, boat, or wrench, for all kinds of measurements such as mass, drag, and drag. stress tolerance. And then he can create dozens of designs of that object, each tailored to different top efficiencies.
In other words, it removes the iteration of the design process and could be applied to the design and engineering of consumer goods and industrial parts, replacing some of the human guesswork in product design and increasing the intuition of the designers themselves.
“A fundamental limitation of typical design optimization techniques is that they require a single objective function to evaluate performance. In most applications, however, several criteria are used to assess the quality of a design, ”the document explains. “The structures must be stable and light. Vehicles must be aerodynamic, durable and inexpensive to produce. In most cases, the performance objectives are not only multiple but also conflicting: improving a design on one axis often decreases its quality on another axis. In reality, designers and engineers navigate a complex landscape of trade-offs, generating objects that may not optimize a single measure of quality, but rather are considered optimal in a given performance trade-off.
Take the humble key. Why is he the shape he is, with that bulbous head and a lean neck? This is exactly what MIT’s system can respond to as it generates alternatives to the key we all know by redesigning the design to optimize it for a myriad of factors at once. Through its generated CAD models, we see that a wrench with the most torque for strength would actually have a very elongated and thin design. But there’s a catch: it couldn’t take a lot of stress. A short, big wrench can take a lot of stress, but it can’t generate a lot of torque. And in fact, in the balance of all the most important factors – weight, torque and strain – we see the wrench design we all know.
In other words, this machine logic can do in seconds what people have perfected over the centuries, arriving at the exact same conclusion, which is both valid and demeaning.
But of course, we know how to make an adjustable wrench, and we’ve known that for a long time. What’s so promising about MIT’s research is that it can run on virtually any CAD model you throw at it, and for the most part, do so within existing workflows. This means it could help designers optimize their existing processes and, most importantly, deconstruct what works and what doesn’t earlier.
Right now, all of these design alternatives are spat out through intricate graphics that are difficult to navigate for the average person, and the software is also not available in any kind of downloadable tool that you can run. (The video you see above is from a different project from last year.) The researchers recognize this: while they’ve created an AI that could replace a team of designers, make it work fine for those designers. is a whole different challenge: “An important consideration comes from human-machine interaction… what is the best way to display it to an engineer who has to digest the space of candidate designs? They ask in the conclusion of the article.
Could we become too reliant on a single system with just one design optimization approach, and perhaps lose some of our own creativity and ingenuity in the process? Indeed, how exactly AI works with the designers of tomorrow is the billion dollar question behind the future of design, let alone the world’s next big key.