{"slug": "mangoguard-edge-ai-that-detects-mango-diseases-in-the-field", "title": "MangoGuard — Edge AI That Detects Mango Diseases in the Field", "summary": "A developer built MangoGuard, an edge AI system that detects mango diseases directly on an Arduino Nano microcontroller without requiring cloud connectivity or internet access. The quantized MobileNetV1 model classifies mango leaf disease in under two seconds with 86.45% accuracy on real Ethiopian farm data, while a Raspberry Pi gateway evaluates environmental disease risk and generates plain-language recommendations. The system streams results to a bilingual English and Amharic React dashboard via WebSocket, with all scan data stored in PostgreSQL for use as a labeled dataset.", "body_md": "*Submission for the GitHub Finish-Up-A-Thon Challenge*\n\nMango farming is a lifeline for millions of smallholder farmers in Ethiopia.\n\nA single fungal outbreak can silently destroy 20–30% of a harvest before a\n\nfarmer even recognises it. Existing solutions require a lab, a specialist,\n\nor reliable internet. None of those exist where the problem is worst.\n\n**MangoGuard runs AI directly on a microcontroller smaller than a credit card\n— no cloud, no Wi-Fi, no lab.**\n\nAn **Arduino Nano 33 BLE Sense** runs a quantized MobileNetV1 model that\n\nclassifies mango leaf disease in under 2 seconds at 86.45% accuracy on real\n\nEthiopian farm data. The Nano also reads live temperature and humidity via a\n\nDHT22 sensor and sends everything — disease result, temp, and humidity — to a\n\n**Raspberry Pi 4 gateway**, which evaluates environmental disease risk against\n\nagronomic thresholds, runs a 24-hour AI forecast model, and **generates\nplain-language recommendations** pushed directly to farmers based on current\n\nEverything streams to a React dashboard in real time via WebSocket.\n\nFrom the dashboard you can:\n\nThe dashboard is fully bilingual — **English and Amharic (አማርኛ)** — because\n\nagronomists advising Ethiopian farmers shouldn't have to work in a language\n\nthat isn't theirs.\n\nUnder the hood, every scan is saved to a PostgreSQL database. The **admin\ndashboard** exposes this data as a labeled dataset that can be used to\n\n**🌐 Live Demo:** [https://mango-guard.vercel.app/](https://mango-guard.vercel.app/)\n\n**📂 GitHub:** [https://github.com/SCIFI-Shinobi/Intelligent-Mango-Health-Monitoring](https://github.com/SCIFI-Shinobi/Intelligent-Mango-Health-Monitoring)\n\nThe hackathon prototype worked — but only on my machine, with undocumented\n\nsecrets, and no way for anyone else to run it. Here's what I shipped to fix that:\n\n`ARCHITECTURE.md`\n\n, `DEPLOYMENT.md`\n\n, and `CONTRIBUTING.md`\n\n`.gitignore`\n\nwhich was blocking `.env.example`\n\nfiles from being committed`scanIntervalMs = 3600000UL`\n\n, forecast threshold ≥ 24.A stranger can now fork, configure, and deploy this in under 20 minutes.\n\nCopilot saved the most time on the backend — `main.py`\n\ngrew past 3,000 lines\n\nand it was excellent at continuing repetitive patterns: email templates,\n\ndatabase migration helpers, route structure. Once I wrote the first, it nailed\n\nthe second.\n\nDuring the polish phase, asking Copilot to review the README flagged that I\n\nhad no troubleshooting section and no env variable docs — exactly what a\n\nfirst-time contributor needs. It also helped clean up firmware comments after\n\nI fixed the production bug.\n\nThe honest limitation: it continues patterns well but won't proactively tell\n\nyou what's missing. You have to ask the right question first.\n\n*Built to bridge the gap between AI and smallholder agriculture in Ethiopia. 🌱*", "url": "https://wpnews.pro/news/mangoguard-edge-ai-that-detects-mango-diseases-in-the-field", "canonical_source": "https://dev.to/eyobel_z/mangoguard-edge-ai-that-detects-mango-diseases-in-the-field-h93", "published_at": "2026-06-03 03:53:06+00:00", "updated_at": "2026-06-03 04:12:11.341944+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "computer-vision", "ai-products", "ai-startups"], "entities": ["MangoGuard", "Arduino Nano 33 BLE Sense", "MobileNetV1", "Raspberry Pi 4", "DHT22", "React", "PostgreSQL", "GitHub"], "alternates": {"html": "https://wpnews.pro/news/mangoguard-edge-ai-that-detects-mango-diseases-in-the-field", "markdown": "https://wpnews.pro/news/mangoguard-edge-ai-that-detects-mango-diseases-in-the-field.md", "text": "https://wpnews.pro/news/mangoguard-edge-ai-that-detects-mango-diseases-in-the-field.txt", "jsonld": "https://wpnews.pro/news/mangoguard-edge-ai-that-detects-mango-diseases-in-the-field.jsonld"}}