Monowheel Engineer: GPT-6 Astra and V2Fun 3D Workflow
A seated engineer meets procedural machinery in this hybrid 3D case. GPT-6 Astra plans and coordinates the assembly, V2Fun generates the character, and Three.js supplies the monowheel structure.
Interactive 3D preview · Monowheel Engineer: GPT-6 Astra and V2Fun 3D WorkflowDrag to rotate · Scroll to zoom
Take a Closer Look
-
01 / InputInput · Monowheel Engineer: GPT-6 Astra and V2Fun 3D Workflow
-
02 / ProcessProcess · Monowheel Engineer: GPT-6 Astra and V2Fun 3D Workflow
-
03 / FinalFinal · Monowheel Engineer: GPT-6 Astra and V2Fun 3D Workflow
Overview
Monowheel Engineer combines one seated adult woman with a large ring-shaped vehicle. Blue work overalls and dark protective boots distinguish the character, while orange supports, blue-gray crossmembers, and a thick black tire define the machine.
The case demonstrates a division between organic character generation and regular mechanical construction. The technical notes assign component planning and assembly coordination to GPT-6 Astra. The process image identifies the textured character as V2Fun-generated 3D and the mechanical groups as procedural Three.js work. This follows the broader approach described in GPT-6 Astra and V2Fun model collaboration.
Tools Used
- GPT-6 Astra: subject hierarchy, component planning, and assembly calibration, as described in the case notes.
- V2Fun: generated modeling references and the textured seated character.
- Three.js: procedural frame, wheel assembly, seat, mechanical equipment, controls, and foot supports.
- GLB: the supplied final asset format.
Step 1 — Establish the Modeling References
The input board presents a complete vehicle concept, an isolated seated character, and an inferred front, side, and rear study. The isolated reference exposes the bent legs, boots, forearms, and grip pose without the surrounding wheel obscuring them.
The additional views are labeled as AI-inferred. They provide an interpretation of the concept rather than independent observations of hidden geometry. The single-image and multi-view input guide provides related guidance on reference coverage and occlusion.
Step 2 — Divide the Character and Machinery
The exploded presentation organizes the model into six display groups:
- Character: the seated engineer, clothing, and boots.
- Structural frame: orange curved supports, the central platform structure, and blue-gray crossmembers.
- Running gear: the large tire and concentric inner rings.
- Body and seat: the dark seat cushion and backrest.
- Mechanical equipment: a rectangular equipment box and cylindrical hardware.
- Controls and supports: paired control assemblies and foot platforms.
These are presentation groups, not a verified mesh count. Their separation makes the relationship between the generated figure and the procedural machine visible.
Step 3 — Generate the Character and Build the Machine
V2Fun supplies the complex character form. The process and final images show facial features, folded blue clothing, and dark boots on a figure already in a seated pose. The relevant generation route is covered by V2Fun's image-to-3D model generator.
Three.js supplies the regular mechanical forms shown around that figure: concentric rings, repeated tread blocks, curved orange members, rectangular supports, and cylindrical joints and rollers. The evidence supports this hybrid construction method without establishing a generation-time or token-saving benchmark.
Step 4 — Assemble Around the Seated Pose
The final view places the seat and platform inside the wheel, with support arms extending toward circular side joints. The character's hands meet the short control grips, and both boots rest on the forward foot platforms. The equipment box sits beneath the seating area.
These visible contact relationships are central to the composition. The final presentation identifies the image as a render of the delivered GLB after reloading, with unseen surfaces inferred. It documents a static display assembly.
Workflow
Concept and isolated character references → component planning → V2Fun character generation and Three.js mechanical construction → assembly calibration → GLB export → documented reload and final render.
Prompt
English translation of the design intent:
An adult female engineer with curly hair, wearing blue work overalls and protective boots, sits inside a large monowheel balancing vehicle with both hands holding short joysticks. The machine has a thick, large-diameter tire surrounding concentric metal rings. A horizontal seat platform is suspended centrally and carries a seat, battery box, and footrests. Side support arms connect to central bearings. Her head and shoulders extend above the upper edge of the wheel ring, and her face remains unobstructed. Use blue-gray metal with orange structural parts.
Create one complete standalone asset containing one character and one machine. Combine realistic proportions with mild exaggeration in a collectible model style, with an expressive face and natural clothing folds. Build the machinery from clear cylinders, discs, rectangular plates, straight rods, and a few regular curved tubes. Keep major part boundaries distinct, connections plausible, and contact between the character, seat, grips, and footrests intact.
Show a front three-quarter view from slightly above, with the entire subject centered and space around it. Use a white background, soft studio lighting, contact shadows, clear focus across all parts, and refined PBR materials. Exclude text, numbers, logos, watermarks, complex scenery, and motion blur. Produce one finished reference image, not a collage or exploded view.
Result
The final render shows a complete seated engineer and monowheel assembly, retaining the blue clothing, orange frame, dark tire, paired controls, and supported boots. The result differs from the written brief in one visible proportion: the engineer's head remains below the top of the wheel rather than projecting above it. The rendered hairstyle is also more compact than the loose curls in the character reference.
Supplied output metadata:
- Filename:
80-v3-final.glb - Size: 44,656,108 bytes, approximately 44.66 MB.
- MIME type:
model/gltf-binary.
The documented outcome is a static display model. Material-map specifications and mechanical balancing behavior are not established by the images. For a subsequent file handoff, the FBX, GLB, OBJ, and USDZ export guide explains format selection and destination checks.
Frequently Asked Questions
Which part of the monowheel engineer was generated with V2Fun?
The process board identifies the textured seated character as V2Fun-generated 3D. It labels the five mechanical display groups as procedural Three.js work.
How was the monowheel assembly divided?
The exploded view shows six groups: character, structural frame, running gear, body and seat, mechanical equipment, and controls and supports. These groups do not establish the GLB's exact mesh count.
How does the character fit the vehicle in the final render?
The engineer sits within the wheel on a central seat, with hands at the paired short grips and boots on the forward foot platforms. The frame and side joints surround this seated pose.
Does the final model match every detail of the prompt?
It preserves the main subject, colors, seating arrangement, and monowheel silhouette. However, the head remains below the wheel's upper edge, and the hairstyle is more compact than the curly reference.
What file was delivered, and does it demonstrate a working balancing vehicle?
The supplied metadata lists 80-v3-final.glb at 44,656,108 bytes with MIME type model/gltf-binary. The final presentation describes a static display model; it does not demonstrate balancing simulation or vehicle motion.
Explore More
-
GPT-6 Astra x V2Fun - From Code Generation to Model Collaboration | V2Fun
See how GPT-6 Astra and V2Fun combine agent reasoning, code generation, and AI 3D foundation models to build complex, structured 3D assets.
View page → -
AI 3D Model Generator Input Guide: Single vs. Multi-View | V2Fun
Learn how to prepare photos, multi-view references, and sketches for an AI 3D Model Generator, then inspect geometry and textures before export.
View page → -
V2Fun - AI 3D Model Generator
Create 3D models from text prompts or images with V2Fun's AI 3D model generator. Generate characters, props, game assets, and 3D-printable models faster.
View page → -
AI 3D Model Generator Export Guide: FBX, GLB, OBJ, or USDZ? | V2Fun
Use an AI 3D Model Generator with the right export format. Compare FBX, GLB, OBJ, and USDZ for Blender, game engines, animation, web, and AR.
View page →