Skip to content

PROPOSAL_FEATURE_ENHANCEMENTS.md

Latest commit

 

History

History
116 lines (101 loc) · 6.51 KB

PROPOSAL_FEATURE_ENHANCEMENTS.md

File metadata and controls

116 lines (101 loc) · 6.51 KB
Raw

Proposal: New Feature Enhancements for AgriTech

Overview

  • Purpose: Add four major features to AgriTech to increase farmer adoption and provide actionable, expert-backed guidance.
  • Features:
    1. AI Fertilizer Recommendation System
    2. Weather Forecast Dashboard Integration
    3. Expert Q&A Forum (Community Upgrade)
    4. Farm Progress Tracker

Scope & Goals

  • Deliver production-ready backend APIs, minimal frontend components, and integration points for ML and third-party APIs.
  • Keep changes incremental and backward-compatible with existing crop_recommendation module and dashboard.html.
  1. AI Fertilizer Recommendation System
  • Goal: Suggest fertilizer types and application rates based on soil pH, crop type, and weather conditions.
  • Inputs: soil_pH, crop_type, growth_stage (optional), recent_weather (temp, rainfall).
  • Output: fertilizer recommendation object {fertilizer_name, NPK_ratio, application_rate, notes}
  • Implementation approach:
    • Train a lightweight ML model (scikit-learn or TensorFlow/Keras) using historical fertilizer-response datasets.
    • Integrate model inference into crop_recommendation module (e.g., new function recommend_fertilizer()).
    • Expose API route: GET /api/recommend/fertilizer (query params or POST JSON payload).
    • Validate inputs and return confidence score.
    • Model storage: add /models/fertilizer_model.* and versioning metadata in ipm_config.json or ml-model-versioning-implementation.md.
  • Testing: unit tests for inference wrapper, integration tests for API route.
  1. Weather Forecast Dashboard Integration
  • Goal: Provide real-time weather updates on dashboard using OpenWeatherMap API.
  • Backend:
    • New API route: GET /api/weather?lat={lat}&lon={lon} — server fetches OpenWeatherMap 5-day/3-hour forecast and caches for short TTL (e.g., 10 minutes).
    • Config: use env var OPENWEATHER_API_KEY and WEATHER_CACHE_TTL.
    • Rate limiting & error handling for API failures.
  • Frontend:
    • Add a weather widget to dashboard.html (or dashboard component) showing current conditions, 3-day summary, and an icon set.
    • Use existing index.js/dashboard.js pattern. Minimal UI with optional Chart.js sparkline for temp/precip.
  1. Expert Q&A Forum (Community Upgrade)
  • Goal: Allow verified experts to answer farmer queries, with role-based access control.
  • Auth & Roles:
    • Add roles: admin, farmer, expert.
    • Use JWT-based authentication. Extend current auth_utils.py to support roles in token claims and role-check decorators.
    • Endpoints:
      • POST /api/questions — create question (farmer)
      • GET /api/questions/:id — view question and answers
      • POST /api/questions/:id/answers — add answer (expert only)
      • GET /api/questions — list (paginated)
    • Verification flow: admin can mark a user as expert via PUT /api/users/:id/role or via DB flag.
  • DB changes:
    • questions table: id, user_id, title, body, tags, created_at, status
    • answers table: id, question_id, user_id, body, created_at, upvotes
  • Frontend:
    • Update forum UI pages (community_forum.html or forum.js) to show expert answers with badge and allow filtering by expert.
  1. Farm Progress Tracker
  • Goal: Farmers log crop activities and view visual progress insights.
  • Data model:
    • activities table: id, user_id, farm_id, crop_id, activity_type (sowing, irrigation, fertilizing, harvesting), date, notes, attachments
    • crop_progress aggregated view/table to compute GDDs, days since sowing, expected harvest window.
  • API:
    • POST /api/activities — log activity
    • GET /api/activities?user_id=&farm_id=&crop_id= — list activities
    • GET /api/activities/summary — summarized metrics for charts
  • Frontend:
    • New farm_progress.html or widget for dashboard.html with charts (Chart.js) showing activity timeline, irrigation frequency, and expected harvest.

Security & Auth

  • Extend JWT tokens to include role claim.
  • Use auth_utils.py to implement @requires_role('expert') and @requires_role('admin') decorators.
  • Store sensitive keys in environment variables; do not commit API keys.

Data & Schema Migration

  • Add migration scripts (SQL or a small migration helper) to create roles, questions, answers, activities tables.
  • Backward-compatible migrations: add new columns with defaults where possible.

Integration Points & Files to Modify

  • Backend:
    • app.py — register new blueprints/routes for /api/recommend/fertilizer, /api/weather, /api/questions, /api/activities.
    • auth_utils.py — JWT role support and decorators.
    • agri_utils.py or crop_recommendation module — add recommend_fertilizer() wrapper and model loader.
  • Frontend:
    • dashboard.html and dashboard.js (or farm_dashboard.html / farm_dashboard.js) — add weather and progress widgets.
    • community_forum.html and forum.js — expert Q&A UI updates.
    • Add Chart.js to index.html or per-page scripts (use CDN in prototype).

ML Model Notes

  • Prototype with scikit-learn (RandomForest) or a small neural net with Keras depending on dataset size.
  • Provide training notebook in /ml/ with dataset preprocessing steps, reproducible training and eval.
  • Maintain model metadata: version, training_date, metrics.

Timeline & Milestones (Suggested)

  • Week 1: Design data model, auth roles, and API contracts; create migrations.
  • Week 2: Implement JWT roles and forum endpoints; basic frontend for Q&A.
  • Week 3: Implement activities API and farm progress UI; basic charts.
  • Week 4: Integrate OpenWeatherMap API; add weather widget and caching.
  • Week 5: Train POC fertilizer model, integrate into crop_recommendation and expose API; testing and polish.

Testing & QA

  • Unit tests for new API routes and auth decorators.
  • Integration tests for end-to-end flows (question->answer, activity->chart)
  • ML model CI: add smoke test ensuring model loads and returns expected output format.

Deliverables

  • PROPOSAL_FEATURE_ENHANCEMENTS.md (this file)
  • New API routes registered in app.py and helper modules
  • Minimal frontend components wired into existing pages
  • Migration scripts and sample data
  • ML training notebook and model artifact

Open Questions / Decisions

  • Choice of ML framework (scikit-learn vs TensorFlow) — dataset dependent.
  • Whether to persist cached weather in DB or memory cache (Redis recommended for scale).
  • Expert verification workflow (manual admin vs third-party verification).

Next Steps

  • Review proposal and approve scope.
  • I can scaffold the API routes, add DB migrations, and scaffold frontend widgets next — which would you like me to start with?