Empowering aquaculture through automation, AI, and real-time insights.
FIRMATECH is an AI-integrated, modular aquaculture platform built for real-world farm automation. It supports intelligent fish counting, disease detection, biomass estimation, environmental monitoring, and more.
- Real-time tracking of critical water parameters:
- Oxygen (O₂)
- pH level
- Temperature
- Nitrite / Nitrate
- Ammonia (optional)
- Rule-based threshold detection with customizable limits
- Smart notifications (Push/SMS) when thresholds are exceeded
- AI-based turbidity (water clarity) estimation via camera
- Feeding history management
- Local data caching with automatic cloud synchronization
-
Instant anomaly detection via sensor or image analysis
-
Alert types:
- Critical — Push + SMS
- Warning — Push only
-
Real-time alert feed with resolution tracking
- Fish Counting — Real-time or periodic
- Biomass Estimation — Length prediction model
- Disease Detection — For Tilapia (e.g., lesions, discoloration, abnormal behavior)
- Simple interface for low-literacy, rural users
-
End-of-day AI-generated reports:
- Health status
- Water quality summary
- Growth trends
- Feed efficiency (if applicable)
- Suggested actions
-
Export options:
- PDF for digital
| Purpose | Repository Name |
|---|---|
| Frontend Dashboard | firmatech-dashboard |
| Mobile/Web Application | firmatech-app |
| Core Backend (REST API) | firmatech-backend |
| AI & Inference Service | firmatech-backend-ai |
| Faker Data & Simulators | firmatech-fake-data-simulation |
FIRMATECH consists of interconnected backend services for sensor/image/video simulation, real-time processing, AI inference, and data management.
- Sends simulated sensor readings to the backend at fixed intervals
- Backend checks threshold rules
- Triggers alert if abnormal values are detected
- Mimics real-time pond monitoring
- Sends a simulated fish image to the AI backend on schedule
- AI model estimates fish weight
- Result is used to predict harvest readiness
- Represents snapshot-based fish monitoring
- Simulates tank video feed
- AI backend detects and counts fish from video
- Useful for estimating fish density and real-time activity
- AI model is ready; video feed simulation is under development
- Core API services
- Dataset CRUD operations
- User & role management
- Interfaces with frontend, mobile app, and other services
-
Runs at the end of each day
-
Collects:
- Sensor metrics
- Feeding records
- Fish count or disease status (if available)
-
Sends data to the Agentic AI service
-
Receives back:
- A health/growth summary
- Recommendations
- Report for export (PDF/SMS)
