StrideQuest: Touch Grass, With a Little Help From On-Device AI

Description and use

StrideQuest is a local‑first PWA that turns a walk into a mini‑expedition, using in‑browser AI to classify natural objects and storing data locally to encourage real‑world exploration.

Problem

Users spend excessive screen time; need an app that encourages outdoor activity while still leveraging AI for discovery.

Outcome

StrideQuest demonstrates that open‑weight models and web standards can power a privacy‑first, local‑first outdoor scavenger hunt that rewards physical exploration instead of screen time.

Workflow

• Select a quest from the dashboard
• Step outside and pause the phone to photograph a target item
• Use the camera viewfinder to capture an image
• The in‑browser AI model classifies the image and suggests matches
• Save the observation to the local field journal
• Earn XP and unlock badges as you complete expeditions
• Review expedition history and telemetry offline

Evidence

StrideQuest is a local-first outdoor scavenger-hunt PWA for the DEV Hacktoberfest Week 1 Touch Grass challenge.

Source

All field observations, geographic metrics, and raw image blobs are stored client-side in IndexedDB, while interface configurations live in localStorage.

Source

The vision pipeline runs entirely in-browser using Transformers.js v3 and ONNX Runtime Web.

Source

The model is a generalized ImageNet-1K classifier, not a specialized taxonomic or botanical authority.

Source