Ranks Bhopal’s Visarjan ghats for water-quality monitoring from historical MPPCB readings, each traced to its report page, with Hindi or English briefings.
The problem
Water-quality testing reports for Bhopal’s Visarjan ghats contain important environmental information, but the results are buried in lengthy PDF documents. Municipal officials must manually compare readings across locations and years to identify concerning conditions and decide where monitoring and maintenance should be prioritised.
Ghat Watch addresses this problem by turning historical water-quality data into an evidence-based, easy-to-understand decision-support tool. It helps officials identify priority ghats, investigate water-quality trends, trace findings to source reports, and recognise missing data instead of mistaking a lack of evidence for clean water.
How it works
Ghat Watch is a bilingual environmental decision-support platform designed to help prioritise water-quality monitoring at Bhopal’s Visarjan ghats.
It transforms historical MPPCB water-quality readings into a ranked dashboard using transparent, configurable scoring rules based on relevant CPCB Class B benchmarks. Users can explore dissolved oxygen, BOD and pH readings, compare conditions before, during and after immersion, inspect individual ghats, and trace readings back to the original report pages.
The platform also provides proposed action plans for different stages of immersion, ranking sensitivity analysis, data-quality warnings, and one-page Hindi or English briefings for decision-makers.
Our focus is on making environmental evidence actionable, traceable and transparent about its limitations. The current prototype uses historical data and does not claim to provide live water-quality measurements or predict future conditions.
How Claude was used
We used Claude and Claude Code as development collaborators to analyse the problem, structure the application, implement and refine features, and improve the reliability of the codebase.
Claude Code helped us work with the existing repository, develop the data-processing and scoring workflow, build the dashboard and decision-support features, and test the application while preserving traceability to the original water-quality reports.
We used AI assistance to accelerate development, reason through implementation choices, and identify edge cases such as missing readings, below-detection-limit values, and uncertainty in historical rankings. We reviewed the resulting code and treated deterministic calculations and source data as authoritative rather than relying on AI-generated environmental claims.
Team
SWAT