AI 3D Model Generation: Revolutionizing 3D Printing?

3D printing has revolutionized product customization and prototyping. Slicers, filaments, and resins are more user-friendly. Though not quite at the 'one-click print' stage, AI's role in design promises to bridge the gap, making 3D printing accessible to everyone. The integration of AI into 3D modeling could solve customization and accelerate 3D printing adoption.

Accessibility and Ease of Use

3D printing has evolved significantly, becoming easier to use for both hobbyists and professionals.

Slicers are now more optimized, and the materials used—filaments and resins—are less finicky. These improvements have reduced the learning curve, inviting more people to explore 3D printing. While printers themselves are becoming more sophisticated and automated, some challenges remain before it becomes an entirely seamless experience.

Despite these advances, the 'one-click print' dream remains elusive. Technical hurdles still exist, with potential issues ranging from bad settings and leveling problems to maintenance-related issues that can plague even the most advanced 3D printers. But the biggest obstacle is enabling easy customization.

The Customization Conundrum

One of the biggest promises of 3D printing is easy customization of products and prototypes.

3D printers should empower users to design unique and tailored items. However, most users don’t create their own designs but rather rely on pre-existing models from online repositories. While the convenience of these models is appealing, they often require adjustments to fit specific needs, necessitating knowledge of CAD software.

For those new to 3D printing, learning CAD can be intimidating. This challenge slows down the process and limits the ability to customize. Companies are trying to bridge this gap by integrating model hosting sites directly into their slicers. This approach aims to streamline the workflow, allowing users to import models with optimized settings for printing.

AI Image Generation: A Glimpse of What’s to Come

Artificial intelligence has made notable advancements, particularly in image generation. AI image generators such as Midjourney

create incredibly detailed images from simple text prompts. These tools showcase AI's potential to generate complex designs quickly, although accuracy remains an issue. What if this AI capability were applied to 3D model generation?

The concept involves creating a vast data set by rendering a subject 200 times from different angles, using that data set, import to software and create a 3d model. By generating models based on AI, 3D modeling and printing would take a revolutionary step.

Text-to-3D: Testing the Waters

It is now possible to have AI create 3D models based on text.

Meshry, Alpha3D, Sloyd, and Genie are free-ish tools that create 3D models from prompts, but there is still a lot to be improved.

• Meshry: Can be seen as a good platform, but does not generate hollow objects and there are issues with the colors -Alpha3D: Makes low-poly, vague looking, but sometimes okay 3D models. -Sloyd: It uses a pre-made model. It can be said that it doesn’t really create objects. The model has to be already in the catalog, to begin with. -Genie: Heavy on textures to provide depth, the overall quality leaves a lot to be improved as well.

To give a better look into the qualities of the platforms, here is a table with the most important aspects

3D modeling service qualities

As can be seen in the list, the text-to-3D technology still has a long way to go.

Limitations and Opportunities

Although AI in 3D design can be a useful and rapid way to design 3D models, it has a lot of downsides. The primary issues is that models create with AI cannot directly be used in 3D prints. This limitation means it can only be used as a starting point. The potential here is to create a model and import it into sculpting or CAD software to perform additional geometry changes to make it suitable for printing. The AI can speed up the geometry creation, the printing process is still manual.

The Path to True One-Click Printing

The discussed AI Tools were disappointing due to them not being optimized for 3D printing . They create low-resolution geometry and are highly reliant on textures to impart depth. However, generating textures with AI that can be imported in sculpting software such as blender can be usefull, allowing for starter models for 3D printing, therefore it can provide customization. While AI-generated 3D models aren’t ready for one-click printing, these tools could give a head start with 3D models, but more details are needed.

In the future, AI image generation didn’t even exist for a couple of years, so the possibilities that may open in the next years are still unknown.