LankaLens — AI Nature Explorer

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built
🌿 LankaLens — AI Nature Explorer

I built LankaLens, a mobile application that encourages people to put their phones down and explore the nature around them.

The idea is simple: when you're walking, hiking, gardening, or exploring outdoors, you can take a photo of a plant, tree, flower, insect, or other part of nature. LankaLens uses open-source AI to help identify and explain what you found.

But the goal isn't to keep people looking at the app.

The app gives you a short explanation and an observation challenge, such as:

"Look around you and see if you can find another tree with similar leaves."

Then the phone goes back into your pocket.

I designed it for people who are curious about nature but don't necessarily know the names of the things they see around them.

The longer-term goal is to make it particularly useful for exploring Sri Lankan plants, trees, birds, insects, and local biodiversity, with support for English, Sinhala, and Tamil.

Demo

🎥 Demo video: [Coming soon]

The demo will show the application being used outdoors with the device disconnected from the internet, demonstrating the local-first AI workflow.

Code

💻 GitHub: [Coming soon]

The project will be open source so others can experiment with the models, improve the identification pipeline, add local species knowledge, and adapt it to their own regions.

How I Built It

The application is built around the idea that AI should work for the outdoor experience rather than become the experience.

The mobile application handles:

📷 Camera-based nature observations
📍 Location-aware exploration
💾 Offline storage
🌿 Nature observations and history
🧭 Small outdoor exploration challenges
🌐 English, Sinhala, and Tamil support

At the core is an open-weight AI model running locally rather than sending every photograph to a proprietary cloud AI API.

The architecture is designed so that the AI model can be replaced or upgraded without rebuilding the entire application.

         Mobile App
             │
    ┌────────┼────────┐
    │        │        │
 Camera     GPS    Offline DB
    │
    ▼

Local AI Inference
│
▼
Open-Weight Model
│
▼
Nature Identification
│
▼
Short Explanation
│
▼
Exploration Challenge
│
▼
📱 → Pocket
🌳 → Explore
Why Does Open Innovation Matter?

For this project, open AI isn't just a technology choice. It changes what the application can do.

Nature exploration often happens in places where mobile connectivity is unreliable or unavailable. A cloud-only AI application would make the experience dependent on an internet connection and an external service.

With an open-weight model and local inference, the goal is to make the core experience possible without an internet connection.

It also means users don't have to upload every photograph they take to a third-party AI service.

More importantly, an open approach gives the project room to evolve.

I can experiment with different models, improve the prompts and inference pipeline, add local knowledge about Sri Lankan species, and eventually fine-tune or replace components without being locked into a single proprietary API.

For this project, open innovation makes AI more accessible, private, customizable, and useful in the places where people actually go outside.

My Agent Session

Coming soon.

I will include the DevRelay session showing how the project was designed and built.

Prize Categories
Open-Source AI
Touch Grass