AI-Driven Generative Design Redefines the Engineering Process
Image courtesy of Inhabitat
Today, artificial intelligence (AI) and machine learning (ML)-based systems enable applications ranging from junk mail or text filters to autonomous vehicles and robots. Engineering firms and their employees have utilized AI and ML to improve the engineering design process and create highly optimized and original products. Early adopters have benefitted from shorter design cycles, engineering productivity and originality, which has transformed entire engineering and product development workflows.
Let’s examine where traditional engineering design techniques and processes can be exceeded by generative design. We’ll consider some real-world examples to realize why increasing numbers of engineers are turning to AI and ML techniques to reinvent and optimize the design process.
Current Engineering Design Drawbacks
Standard engineering design techniques are well understood, widely utilized, and broadly applicable. The familiarity with—and success of—traditional approaches has delivered countless engineering breakthroughs that we benefit from every day. The engineering design process consists of several steps:
Idea and conceptual breakthrough—Identify the problem and come up with a concept
Create—Fabricate a rough prototype of the concept (ideally in silico)
Refine design—Fill gaps with a comprehensive design (also ideally in silico)
Computer aided engineering (CAE) Validation—Test the design to establish that it works (again in silico)
Manufacture—Choose optimal production techniques and build the product at scale
Launch—Release the product into the market, ideally with sales and marketing materials generated directly from computer aided design (CAD) files with graphic processing unit (GPU) accelerated photorealistic rendering, virtual reality (VR), and other visually compelling techniques to complement mainstream marketing tools
This process is inherently linear and glosses over significant drawbacks associated with this approach.
Extensive technical expertise is required at each step. Although advanced software is used, every dimension, specification, and feature must be exactly defined using elaborate, domain-specific software tools to realize a practical design that is ready to manufacture. Negative feedback loops occur when something goes wrong during the validation phase, which invariably delays projects. This can result in product recalls, redesigns, and wasted resources. Engineering and design creativity are limited by how quickly teams can iterate and generate new designs. With tight schedules, the normal ‘safe’ approaches reign, which significantly impedes long-term innovation.
Engineering software has ameliorated some of these issues, but engineers still have to complete each development phase. New technologies have the potential to amplify engineering and design productivity. Generative design can increase product development efficiency and utilize new fabrication techniques such as 3D printing (additive manufacturing).
Generative Design Changes Everything
Generative design utilizes AI and ML to transform tedious engineering design processes into a seamless interaction between computer and engineer. Topology optimization and simulation is automatically performed by the computer. Negative feedback loops are removed by lowering barriers to design, giving engineers more room to tackle challenges that require “common sense.”Generative design can optimize a design for specific parameters, such as weight or durability, or commercial parameters, like production costs and aesthetic considerations. Most intriguing is its ability to enhance functionality (by design) during use. Startups such as additive flow have delivered applications that enable engineers to integrate different materials into one component while optimizing the topology in parallel. This results in superior products and shorter development times.Generative design works best in conjunction with other technologies—generative design and 3D printing are a classic example, making it possible to quickly prototype and test new designs without a costly and time-consuming custom manufacturing run. 3D printers have no geometric boundaries, so extremely complex structures can be delivered.3D printing also facilitates mass-customization. It can print products tailored to the specific needs of a single client. Imagine using AI to create a perfect part, just for your product design, without the economic drawbacks inherent in traditional subtractive manufacturing procedures.
Generative Design in Practice
Perhaps you’re designing a motorcycle swingarm. After coming up with a design area, connection points, and constraining parameters such as weight or torque, you spend hours calculating whether a proposed design would meet each of the engineering requirements—and that’s just for a single proposed design.
In addition to saving time, generative design algorithms can unlock entirely new approaches that weren’t feasible before.To understand the generative design advantage, we must take a look at classic topology optimization algorithms. These minimize designated objects according to predetermined constraints, for instance volume or weight. Generative design algorithms utilize these steps but allow a wider range of constraints. Engineers add a greater variety of requirements, essentially a “fitness landscape” that drives design optimization. The workflow harnesses AI to analyze the use of different materials and manufacturing techniques.
Please click HERE to view the original article.