An app that suggests what you can build with the bricks you own. It recognises parts from photos to build an inventory, and once you pick a theme and difficulty the AI generates a 3D build sequence. A physics engine checks it for overlaps and floating pieces, reports the faults and makes the AI redo it, up to three times.
Judges called the premise, what can I build with the bricks I already own, highly relatable, and found the physics-engine verification loop technically interesting. They also credited a consistent finish from problem framing to experience design.
The core design decision is refusing to take the model's output on trust. A build sequence collapses the moment a single step is impossible, so running the result through a physics engine, checking for overlaps and floating pieces and feeding the specific fault back in words is probably what turned a plausible-looking list into something you can actually build. Capping the loop at three passes reads as an operational judgement too, a trade between wait time and quality.
See all winners of 第4回 Agentic AI Hackathon with Google Cloud →
Summaries are written by this site. Project pages include demo videos, screenshots and the team's own write-up (videos may autoplay).