Wildfire Education & Detection Mobile App
An embedded IoT company develops systems for wildfire detection and monitoring that combine field-deployed devices with machine learning and computer vision models, including a YOLO-based model, to identify wildfire risks in real time. It wanted a public-facing mobile app that showcases its technology, offers educational and informational value to the general public, and demonstrates on-device wildfire detection.
Back to archive
Design and implement a deployable React Native mobile app that surfaces educational content and local wildfire-related news, and integrates the company's existing YOLO computer vision model to analyze camera input on-device and demonstrate wildfire detection.
A lightweight, iterative process: discover and align on requirements and a feasible feature subset, then define the app architecture (navigation, state management, modules) and the integration design for content feeds and the camera and YOLO pipeline, identifying key technical risks such as on-device inference performance early. The team then turned an existing no-code-generated React Native UI prototype into a clean, maintainable codebase, implemented the feeds and camera integration with regular check-ins, and tested and hardened before handover.
A deployable React Native app with educational content, local wildfire news consumed from the partner's feeds and APIs (with graceful handling of no-network and failed-fetch states), and a camera “scan” view that runs the pre-trained YOLO model on-device and shows detection results such as bounding boxes, labels, or confidence. Packaged with iOS and Android build configurations, technical documentation, a short user guide, and a final demo to the partner and faculty.