Inspiration
A landscape can be more than scenery: it can invite care. LakeAlive turns a fictional Bengaluru-inspired lakeside park into a small interactive world where the viewer becomes a participant. Its artistic goal is to connect the beauty of a place with the ordinary actions that help people notice and care about it.
What it does
Players explore a procedural 3D landscape at golden hour as a sun-hatted character. They collect shoreline litter, plant native seeds, and clear inlet litter. These actions produce visible changes in the scene and illustrative clarity, habitat, and wildlife indicators. A self-paced guided tour and field journal make the core experience approachable.
The current source also includes a boating jetty, swimming cove, picnic grove, nursery, quiet bird hide, and fictional neighbours. These support exploration and a sense of place. The supplied video demonstrates an earlier version of the core restoration experience; the repository documents the expanded park.
How we built it
The browser experience uses TypeScript, Vite, Three.js, WebGL, and Lucide icons. Geometry and scenery are generated locally, with seeded placement and deterministic action outcomes. HTML controls sit over the 3D world. Keyboard and touch input, reduced-motion settings, optional synthesized ambient sound, and contextual interactions support different ways to explore.
LakeAlive is existing work created before this InnovArt entry and has appeared in other hackathon submissions. This entry does not claim the code was newly created for InnovArt. AI assistance was used in development and submission preparation; third-party tools are credited above and in the repository. The cover is an AI-generated promotional illustration, not an in-game screenshot; the video and code show the actual procedural 3D experience.
Challenges we ran into
A central challenge is keeping the visual metaphor legible without overstating its scientific meaning. The game explicitly labels its indicators as illustrative, compresses ecological time, and makes actions reproducible. The design also balances a full-screen 3D environment with clear controls and a guided route.
The lake, people, ecological changes, and numerical indicators are fictional. There are no measured environmental outcomes, real-world users, partnerships, or live lake data claimed. It is not a scientific prediction or a recommendation for restoring a real lake. The 3D world requires WebGL and does not provide a complete nonvisual equivalent. The source repository includes setup instructions and validation notes.
Accomplishments that we're proud of
LakeAlive combines environmental storytelling, 3D modeling, animation, game interaction, and interface design. Instead of treating environmental information as a separate chart, it uses changes in a fictional landscape as an expressive response to the participant's actions. It fits InnovArt's interest in immersion, audience participation, and learning through art.
Try the current build: Use the public repository's README: install Node.js 22.13+ and run npm ci followed by npm run dev, then open the local URL in a WebGL-capable browser. No account, API key, or paid service is required. The project can be reset to reproduce the core actions.
What we learned
An interactive landscape can communicate through the consequences of a participant's actions, but that expression needs an explicit boundary between artistic simplification and ecological evidence. Reproducible actions help explain the metaphor. Clear controls and a guided route are part of the artwork's usability, while a complete nonvisual alternative still needs further work.
What's next for LakeAlive
Proposed future work includes a fuller nonvisual alternative, playtesting with learners, and a clearer comparison between artistic simplification and real ecological complexity. These are plans, not completed results.
Log in or sign up for Devpost to join the conversation.