Nature Go

This is a story about a new outdoor exploration app called Nature Go. The app aims to encourage people to disconnect from their screens and engage with the natural world. Instead of offering a traditional dashboard or constant checking, Nature Go presents users with screen-free "quests" such as finding a green leaf, photographing tree bark, spotting a pinecone, or discovering a wildflower.

When an explorer goes outside and captures an image of their discovery, the app verifies whether the requested target is present, whether it is physically real and not a screen or indoor plant, and whether the image shows genuine outdoor context. If verified, the discovery is saved to the explorer's journal, the daily streak and badges are updated, and an educational nature fact is read aloud. If the image is deemed invalid, the app explains why and rejects the discovery.

Nature Go includes 20 nature quests, 8 expedition challenges, 5 achievement badges, and a custom quest creation feature. The app also tracks discovery journal entries with photos and confidence scores, maintaining daily streaks and longest-streak records, and allows for the creation of explorer profiles. The app offers a screen-free audio mode for users who prefer not to use the camera or take photos.

The project's core is an open-source AI verification layer built around Gemma, an open-weight model. This model receives the current quest target and the captured image, returning structured verification data that determines the app's response - playing a success or rejection sound, narrating a nature fact, saving the discovery to the journal, updating quest completion progress, updating streaks and authenticity score, and unlocking badges.

The verification policy checks three conditions: target match, physical authenticity, and outdoor context.

The app's architecture includes a frontend built with React 19, TypeScript, Vite, Tailwind CSS, Motion, and Lucide icons, a backend using Node.js, Express, and TypeScript, Firebase Authentication with Google sign-in and explorer login, Firebase Firestore for database storage, and local persistence using Browser local storage. Audio features are powered by the Browser SpeechSynthesis API and Web Audio API.

The open-source approach behind Nature Go allows the project to be transparent, adaptable, and able to be replaced or improved as new open models become available, catering to different communities with diverse needs, languages, accessibility features, safety rules, and local ecological knowledge. The ultimate goal is to help users complete one quest, hear a nature fact, put the phone away, and notice something else nearby, fostering a deeper connection with the physical world.