Gaps in product and process development are often hidden, not as obvious as automation equipment and factory production organization. When we are racing on the track of Industry 4.0, our competitors are improving the design method and optimizing the design parameters step by step.
The Fourth Industrial Revolution will make copying impossible, because the competition of designs will eventually become a competition of models. For example, instead of comparing specific images generated by MD, SF and DALL.E, we now compare models by comparing the results produced by a bunch of different prompts. The design and development of industrial parts has already begun to roll models before the concept of “Industry 4.0” was proposed.
8440-2252 Topology optimization design is not generative design
Over the past 20 years, topology optimization design has developed rapidly, and now almost all major design software companies offer topology optimization modules, which are perfectly embedded in their respective PLM platforms. Depending on the customer’s manufacturing process and yield requirements, the program can also further optimize the results, automatically generating designs for forging, casting, machining or 3D printing. The drawing Angle for casting is easy, and the pouring gate is automatically designed to be no problem. Even so, this does not count as generative design.
There is a theoretical optimal solution for topological optimization, 8440-2252 namely the perfect truss model proposed by A.G.M. Michell as early as 1904. Therefore, the structures obtained by topological optimization are similar to those of Michell truss.
Generative design should be said to be born out of parametric design, and its ancestor is Autodesk. Simply put, it automatically rearranges the combined design parameters to generate new design results. In this process, many results are not good, so you need a screening criterion, which is simulation. Through screening, the parameters used in the good design results are retained, used as genetic genes, and recombined with other design results, so that several new design results will be produced, and the better results will be screened again, so iterative, which is the so-called genetic algorithm.
Finally, multiple design results are generated and functional parameters are listed for comparison. The application of generative design is wider, the design results are more diverse, and it can carry out multi-objective optimization. For example, to design a more efficient cooling flow channel, the left two are the results of parameter optimization and topology optimization, and the right is the optimization results of generative design. If the part size, material, working 8440-2252 conditions have changed, then the design result will also change dramatically. So at this time to “copy” the design of a single part has become meaningless, in this “finite element simulation based generative AI model” with the support of an original enterprise can continuously produce similar designs at zero cost, the design cost is far lower than the cost of reverse (copy) engineering.
Generative AI will lead the fourth Industrial Revolution
The recent stunning performance of Sora generated videos has made all people cry out that the dream world has come, and the design of proteins with generative AI has also made significant progress in the last year, and I believe that the next breakthrough is the design of 8440-2252 new materials, especially organic materials.
For the manufacturing industry, it is not very difficult to produce new design results, the difficulty is how to screen, so ultimately it is simulation. For example, in the following case, the designer lets the AI automatically generate a large number of car pictures, and then converts the floor plan into a point cloud and then into a vector diagram, and then runs the simulation to obtain the results and screen out more possibilities.
AIGC+CAE (Finite Element Simulation) is about to trigger a design revolution in the industrial and manufacturing sectors, with a depth and impact far beyond “Industry 4.0”. This is how the fourth Industrial Revolution brings a new approach to product design.