TrailWhisper — The 100% Offline Nature Guide Built to Turn Off Your Screen

Abhinav shrivas — Thu, 08 Oct 2026 11:52:31 +0000

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

What I Built

Most modern AI applications are designed to maximize screen time, keeping your eyes glued to glass in endless conversation loops. TrailWhisper is designed to turn itself off.

TrailWhisper is a 100% offline edge nature companion and bioacoustic field guide powered by open-source AI. Built for hikers, trail runners, foragers, birders, and anyone experiencing screen fatigue, it puts nature first:

Ultra-Fast Capture (< 2 Seconds): Point your camera or snap an outdoor specimen (flora, fungi, aviary, pollinators).
On-Device Edge Vision: Open-weight Vision Transformers analyze the specimen locally on your device's GPU (WebGPU) in ~50ms without an internet connection.
Auditory Whisper (Eyes-Up Nature Experience): Instead of making you read dense paragraphs on a glowing screen, a natural voice speaks a concise naturalist insight directly into your earphones, with automated markdown symbol stripping for fluid, human-like cadence.
Tactile Sensory Challenges: Each sighting invites you to physically interact with the wild ("Touch the deeply furrowed white oak bark," "Smell the crushed pine needles").
Anti-Screen Sleep Minimizer (Trail Mode): To keep eyes on the canopy, an automatic sleep countdown dims the display into an ultra-dim, battery-saving OLED black mode with a breathing green leaf ring. (Paused by default for review, with a customizable timer: 30s, 60s, 90s, or 3 mins).
Smartwatch GPX & Markdown Export: Trail observations and offline GPS waypoints export directly into a .gpx file loadable into Garmin, Apple Watch, Strava, and AllTrails—allowing you to pack your phone away completely during hikes.

Demo

GitHub Repository: https://github.com/Abhinav-Shrivas/TrailWhisper

How to Run Locally in 60 Seconds:

git clone https://github.com/Abhinav-Shrivas/TrailWhisper.git
cd TrailWhisper
npm install
npm run dev

Open http://localhost:5173 in any modern browser (Chrome, Edge, Brave).

Interactive Testing Highlights:

Curated Trail Presets: Test immediate edge AI classification without leaving your desk using curated specimens (Eastern White Oak, Fly Agaric Mushroom, Perched Songbird, Autumn Maple, Peat Moss).
Live Camera & Photo Upload: Aim your webcam at houseplants or upload your own trail photos (lions, elephants, mushrooms, trees).
Offline Trail Simulator: Toggle the TRAIL OFFLINE badge in the header to verify that the app runs with 0KB network traffic.
Edge AI Engine Settings: Click the CPU chip icon in the header to inspect your hardware acceleration (WebGPU Direct3D 12/Vulkan vs. WASM SIMD CPU), verify on-device privacy badges (0 bytes uploaded), and manage your local Ollama connection.

Code

Abhinav-Shrivas / TrailWhisper

TrailWhisper is a 100% offline edge nature companion and bioacoustic guide powered by open-source AI. It is built for hikers, trail runners, foragers, birders, and anyone seeking respite from screen fatigue.

🌲 TrailWhisper — Offline Edge Nature Guide & Audio Naturalist

"Built with open-source AI at its core to get people off the screen and into the wild."
Touch Grass Challenge — Open Source AI Build

🍃 The Philosophy: "Make the Screen the Shortest Part of the Experience"

Most AI applications are designed to maximize your time looking at a screen. TrailWhisper is designed to turn itself off.

When you encounter an unfamiliar tree, mushroom, bird, or wildflower on a remote trail:

Quick Capture (< 2 seconds): Point your camera or choose a specimen.
Instant Edge AI Inference: Powered 100% on-device by open-weight AI (WebGPU / Transformers.js).
Auditory Whisper: Instead of reading paragraphs on a screen, a natural voice speaks a concise naturalist insight directly into your earphones.
Tactile Sensory Prompt: You are invited to interact with nature physically ("Touch the deeply furrowed bark," "Smell the…

View on GitHub

How I Built It

TrailWhisper is built from the ground up around open-source AI models and frameworks:

1. Google Gemma 2 (`google/gemma-2-2b-it`) — The Naturalist Reasoning Brain

We integrated Google DeepMind's open-weight Gemma 2 as our ecological reasoning and sensory quest engine:

Custom Naturalist Modelfile: We created a specialized Ollama persona compiled directly on top of FROM gemma2:2b (ollama create trailwhisper -f Modelfile) that produces poetic, scientifically grounded ecological relationships formatted specifically for eyes-up hikers.
Live Local Ollama Bridge: Connects seamlessly to a local Ollama instance on localhost:11434 via a dedicated Vite proxy to stream live 2.6B neural reasoning with zero browser CORS barriers.
Zero-Cloud Distilled Field Fallback: When disconnected on a remote backcountry trail without Ollama active, TrailWhisper automatically falls back to an embedded Distilled Gemma 2 Field Cache with zero network dependency.
Terminal Verification Command: npm run test:gemma

2. Vision Transformer (ViT) — Hardware-Accelerated Edge Classifier

Model: Xenova/vit-base-patch16-224 (open weights in ONNX format).
Runtime: @huggingface/transformers running directly in the browser via WebGPU (Direct3D 12 / Vulkan) with WASM SIMD CPU fallback.
Zero Cloud Latency: Performs forward tensor passes in ~50ms without transmitting pixels across the internet.
Terminal Verification Command: npm run test:model

3. Offline Botanical & Wildlife Taxonomy Engine

Pre-compiled taxonomic field knowledge for 50+ forest species pairing neural predictions with safety badges (Edible / Toxic / Protected), habitat notes, and tactile field cues.

4. Anti-Screen UI, GPX Engine & Local Field Journal

Built with Vite, React 19, and a custom Vanilla CSS design system styled with deep forest obsidian tones (#07100b) and bioluminescent accents (#4ade80).
Observations, timestamps, and GPS coordinates are stored exclusively in local IndexedDB/localStorage with 0 external telemetry.
Export engine formats GPS coordinates directly into standard GPX 1.1 XML schema for cross-compatibility with outdoor GPS watches and mapping software.

Why Does Open Innovation Matter?

1. Datacenters Can't Hike With You

Frontier cloud models (GPT-4o, Gemini Cloud) reside in multi-billion dollar datacenters. When you hike into a valley, canyon, or national park, cellular reception vanishes—and cloud AI goes completely dark. If an assistant requires 5 bars of 5G to tell you whether a mushroom is toxic, it fails the outdoors. By running open weights locally on-device via Transformers.js and WebGPU, TrailWhisper operates reliably in Airplane Mode with 0 bytes sent over the wire.

2. 100% Location & Foraging Privacy

Trail paths and foraging spots are personal and sensitive. Commercial closed APIs log GPS waypoints and store photos on external servers. Open-source AI ensures that all coordinates, images, and notes remain strictly on the user's hardware.

3. $0.00 Operating Cost & Zero Cloud Emissions

No subscription fees, no credit card paywalls, no metered API billing, and zero datacenter water or compute emissions.

4. Freedom to Build an "Anti-Screen" App

Commercial AI companies optimize their products for continuous user engagement and endless chat loops. Open-source AI gave us the autonomy to build an AI that does the opposite: answers in 2 seconds, speaks aloud, and intentionally turns off the screen so you can touch grass.

My Agent Session

This project was built and architected in pair-programming collaboration with Antigravity (Google DeepMind's agentic AI assistant). The agent assisted with:

Architecting the anti-screen WebGPU edge pipeline and auto-sleep countdown mechanics.
Integrating Google Gemma 2 open weights, configuring the custom Modelfile, and building the local Ollama bridge.
Developing automated standalone test suites (test-model.js and test-gemma.js) for reproducible CLI verification.
Configuring GitHub Actions CI/CD to ensure continuous model testing and build verification on every commit.

Prize Categories

1. Best Use of Gemma ($200 USD)

Model Used: Google Gemma 2 (google/gemma-2-2b-it / gemma2:2b) by Google DeepMind.
Implementation: Gemma 2 powers both the dynamic ecological reasoning engine and the real-time sensory quest generator. We created a custom Ollama Modelfile (FROM gemma2:2b) with a specialized naturalist prompt and sensory quest format, and coupled it with an offline distilled field knowledge cache for zero-cloud resilience in the backcountry.
Verification: Run npm run test:gemma to test the open-weight reasoning pipeline directly in your terminal.

2. Best Use of GitHub ($100 USD)

Continuous Integration: Implemented an automated GitHub Actions workflow (.github/workflows/ci.yml).
Automated AI Testing: On every push and PR, GitHub Actions automatically spins up an environment, runs model unit tests to verify the Gemma 2 reasoning engine, and compiles the production bundle with npm run build to prevent regressions for open-source contributors.

DEV Community: Abhinav shrivas