TL;DR
Get business pricing on tools and workshop supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
A maker is using Google’s Gemini image-generation and coding tools to help design and print parts for a Fallout-inspired Halloween prop for his son. The project is still in progress: AI-generated OpenSCAD code has helped produce usable geometry, but the maker reports errors and expects to revise parts or model them manually.
A maker is using Google Gemini to help design parts for a Fallout-inspired Halloween prop for his son, turning some AI-generated OpenSCAD code into test prints while correcting design mistakes. The project is still underway, and the maker says the tool is helping with parts of a process he finds difficult: shaping an idea into components that can be modeled and assembled.
The project began after the maker’s son chose a costume inspired by the video game Fallout 4. A full do-it-yourself power-armor costume was not feasible within their timeframe, so the maker shifted toward a simpler outfit with a decorated chest piece and a distinctive hand-built prop. A planned motorcycle chest protector did not arrive as expected, prompting a possible move to foam and other craft materials.
For the prop, the maker used Gemini to help explore and generate OpenSCAD code. One early design was a truncated cone with a mounting ring. After several iterations, the maker settled on connecting two truncated cones. He says the resulting print, made with filament that was too wet, was good enough to help him assess the design. He plans to enlarge a later version and add mounting holes so it can attach to a wooden box.
The maker also asked the AI to break a fantasy-style tool into printable subassemblies. He found the suggested breakdown unhelpful, although it supported an existing idea to use pins for glued connections. He has edited generated code himself, including changing dimensions and adding holes, and is printing a low-density version of another component. Depending on how that test goes, he may print it in a final color or rebuild the model in conventional 3D software such as Fusion.
AI as a Workshop Design Aid
The account offers a practical example of how generative AI can fit into a small, hands-on maker project: not as a finished-product machine, but as a way to explore shapes and create editable starting points. OpenSCAD code gave the maker parameters he could adjust, which helped him test geometry before committing to a final print.
The limits are just as relevant. The maker reports that Gemini ignored instructions about flush side arms and the placement of pill-shaped holes, and added chamfers he did not want. Those errors meant further prompting, code edits, or possible manual modeling. The project shows how the value of AI in this setting depends on a person checking the output and understanding enough about the design to revise it.
For hobbyists, that distinction matters: a generated model or script may speed up an early step, but it does not remove the work of fitting parts together, selecting materials, testing prints, and making a usable object. Here, the maker is still planning a rough first version, with a wooden box serving as the base for individually designed parts.
FDM 3D printer for DIY prop making
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From Costume Idea to Printed Parts
The maker says he is comfortable modeling objects he has designed before, but finds unfamiliar forms harder when the idea is not yet fully worked out. In this project, the prop’s appearance remained largely in his head, while practical questions—such as how to attach pieces and route visible wires—still needed answers.
Rather than reproduce a complicated tool shown on Fallout-related wikis, he and his son are taking a more creative approach. The prop may become a detector for a synth, ghoul, or mutant, though the maker says his son will decide what to call it. The intended costume includes a hard hat and a flashlight modified to look scavenged. The maker is considering separate printed parts attached to a small wooden box, using screws or adhesive to assemble them.
He has also printed green air fittings using 3D files from McMaster-Carr to see how 1/4-inch-inner-diameter hoses might fit into the assembly. He says this was a way to assess the layout before deciding whether to buy inexpensive metal fittings, rather than ordering an assortment without knowing what would work.
“The AI made a lot of mistakes, and I’m also learning how to better constrain dimensions in my prompts.”
— The maker, in the ToolGuyd report
Design and Assembly Still Pending
The prop has not yet reached a finished version. The maker says he is printing a low-density test part and has not decided whether the final design will stay in OpenSCAD or be remodeled in Fusion or similar software. The final size, color, attachment method, and arrangement of components are also unsettled.
It is not clear whether the rough version will be completed in time for Halloween, or how much of the design will ultimately come from AI-generated code. The source report does not provide independent testing or a comparison of Gemini’s output with other tools. Its account is a maker’s description of one project, not evidence that AI will reliably produce accurate printable assemblies in other situations.
A Rough Version Before Halloween
The maker’s stated plan is to create a rough first-version prop, then refine it if time allows. The next steps include evaluating the current low-density print, adjusting dimensions and mounting features, and deciding how to secure the parts to the wooden base. He may use screws, glue, or other methods to bring the individual pieces together.
He expects to make quick revisions as the design develops and says some parts may be easier to model manually in conventional 3D software. The Halloween deadline is driving the simpler construction plan; any more cohesive redesign may come later. The report does not say whether the prop has been completed.
Key Questions
What is the maker building?
He is developing a Fallout-inspired Halloween prop for his son, likely a fictional detector mounted on a small wooden box. His son will decide what the prop is called.
How is Gemini being used?
The maker has used Google Gemini to help explore designs and generate OpenSCAD code for parts he can adjust and 3D-print. He also asked it to suggest how a fantasy tool might be split into printable subassemblies.
Did the AI produce a finished, accurate design?
No. The maker reports that some instructions were ignored and unwanted features appeared. He has revised generated code and may remodel parts manually; the project is still at the test-print stage.
Is the prop finished?
No completion is reported. The maker plans to assemble a rough first version ahead of Halloween and refine it if time allows.
Source: rss
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
