AI-Powered Tools

AI Water Assistant

Two simple tools to help you analyze water quality and maintain your filter. Upload a photo or enter numbers — get instant recommendations.

Primary Tool

Photo Water Check

Upload a photo of your water sample. The AI analyzes the image to determine turbidity level and provides purification recommendations.

How it works:

  1. The AI analyzes color, clarity, and visible particles in your water photo
  2. It classifies turbidity as Low, Medium, or High based on visual features
  3. You get specific recommendations for settling time and flow rate

Drag & drop water photo here

or

Maintenance Decision Tree

Follow this diagnostic flowchart to determine your filter's maintenance needs

Assessment Start
Evaluate System Performance
Begin by measuring current flow rate
Flow Rate
Normal (>80%)
System operating within parameters
Flow Rate
Low (50-80%)
Reduced performance detected
Flow Rate
Very Low (<50%)
Critical performance degradation
Days Since Clean
<30 Days
Days Since Clean
30-90 Days
Days Since Clean
>90 Days
Fill Time
Normal
Fill Time
Slow (>2x)
Monitor
No Action Required
Normal flow + recent cleaning
Operational
Schedule
Routine Cleaning
Preventive maintenance window
Check: Pre-filter screens, Surface media layer
Maintenance Due
Clean & Inspect
Deep Maintenance
Extended service interval
Inspect: Char media bed, Distribution plate, Under-drain system
Service Required
Repair
System Repair Needed
Performance degradation
Replace: Saturated char media, Clogged screens, Damaged seals
Repair
Urgent Repair
Critical System Failure
Major blockage or structural failure
Check: Inlet pipe blockage, Collapsed media bed, Cracked vessel
URGENT
Monitor
Maintenance
Repair
Urgent

About Our Analysis Tools

Turbidity Analysis

The photo analyzer uses a TurbidityCNN deep learning model built on ResNet18 architecture. It was trained on the Roboflow water-turbidity-ntu dataset to predict NTU (Nephelometric Turbidity Units) values from water sample images. This provides rapid field assessment without laboratory equipment.

Maintenance Decision Tree

The decision tree provides a systematic diagnostic approach based on biosand filter research. By evaluating flow rate changes and time since last cleaning, users can quickly determine the appropriate maintenance action—from routine monitoring to urgent repairs.