Generative design is a process that works within a set of constraints based on which an output is created. The design process is usually undertaken within an algorithm or an application that creates an output based on the set of constraints presented. Based on this output, and as the design goals become more clearly defined, the designer can repeatedly modify the acceptable range of constraints to generate different sets of outputs.
Software solutions integrate this process with the computing power of technologies like artificial intelligence and cloud computing to explore the various design alternatives.
Nature is the first great designer of our world. DNA, which contains the design blueprint of the species, undergoes millions of mutations, each creating a new genetic variant. Out of the millions of variants, only the best, most resilient ones survive the process of natural selection. And thus, our species continues to constantly and automatically improve and evolve.
This same approach, which we see in the evolution of species, is mimicked by generative design. Using generative design we can create hundreds and thousands of potential design options and select the one that is best suited to our requirements.
How Does It Work
Generative design software accepts input parameters that define the problem statement – the goals, requirements, and constraints that the product must satisfy. Based on these inputs, the software generates a potential solution. The designer or the engineer can then evaluate, filter, and fine-tune the generated solution, redefine the parameters and produce the next design to find if it is better suited to the final envisioned product design.
The effectiveness of the design output greatly depends on the accuracy and precision of the input parameters provided. Inputs can cover a wide range of requirements such as geometric dimensions, operating conditions, load-bearing constraints, safety constraints, materials used, manufacturing methods, etc.
The engineer can input as few or as many constraints as they desire. The number and range of output design alternatives would vary accordingly. Although all the alternatives created by the software will meet the requirements specified, they can still vary significantly in shape, materials, manufacturing methods, and cost.
Generative design can offer suggestions on how best to manufacture a part, and even guide engineers towards methods they might not have otherwise considered. When constraints are applied in the form of materials or manufacturing methods, the software only suggests design options that are feasible using the available resources. This allows engineers to compare simulations and analyse the impact of trade-offs between different constraints.
Such simulations are greatly beneficial for designers to figure out the optimum balance between all the constraints. If done manually, creating these designs would be quite an expensive and time-consuming process.
Generative design software is intended to be used as an iterative and adaptive tool. Users can evaluate and rank the output alternatives provided by the software based on how well each solution achieves the pre-defined goals. The interactive process helps the AI-based algorithms to learn the preferred direction of evolution of the design.
Why Use Generative Design?
Generative design greatly simplifies the design process and makes it faster and more streamlined. The series of viable design solutions it offers often serve as the skeleton framework that the engineers can build upon.
Generative design significantly cuts down the amount of time spent ideating. Given a problem statement, instead of starting with a blank drawing board, the designer can use Generative design software to automatically create multiple solutions that they can choose from to work on further.
Generative design also helps the engineers/designers the different ways a particular design problem could be solved. It would be extremely tedious, if not impossible, for an engineer to manually create and compare all possible solutions.
Generative design also helps in exploring solutions that might not have occurred to a designer otherwise. By simultaneously exploring multiple design options and tweaking the priorities of the various parameters, designers can quickly find the ones that strike just the right balance between the cost constraints and the payoff.
This process also brings a significant reduction in manufacturing costs. Design problems are often mind-bogglingly complex. Human mind’s approach to such problems usually involves breaking the singular problem down into sub-parts and creating multiple components that need to be put together later.
With generative design, simple solutions can be created for complex design problems. Especially with the help of additive manufacturing techniques like 3D printing, it is possible to create single components with complex geometries instead of assembling multiple separate parts. This simplifies the supply chain and leads to lower costs.
While generative design does aid the design process significantly, it must not be confused with Computer-Aided Design (CAD). Although CAD has been around for several decades, it has mostly been used only for rendering and optimizing the vision of the designer. The design itself is brainchild of the
designer where the resources required to explore the idea has already been preset without an option of modifying them later.
With generative design, instead of starting a design based on some preliminary ideas, the designer can just tell the computer what they want to build and allow the computer to explore all possible permutations and combinations of the design solution.
This key difference makes design a more collaborative process between the engineer and the software, and results in several advantages while engineering design solutions.
As it is still a new technology area, it is not uncommon for generative design to be confused with other design tools, particularly those that provide optimization functions. This is usually because both tools require similar input parameters. There is, however, significant difference between them.
Optimization tools require an existing design that they then modify to improve its performance against the specified parameters. Generative design, on the other hand, does not require an existing design. Instead, it provides a large number of potential design solutions as output instead of a single optimized solution.
Generative design software is usually tailored to supplement existing CAD platforms. The selected designs can be exported as CAD files which can be further modified and used in downstream processing. They can also be integrated into the wider project.