Satya-Setu

Type: Data Aggregator & Anomaly Detection Dashboard

Satya-Setu is a centralized intelligence dashboard designed to unify fragmented river and pollution data from multiple Indian government and public sources (CPCB, SPCBs, Namami Gange, India-WRIS, data.gov.in, and more). It exposes discrepancies between officially reported data and independent satellite/ground truth, acting as a “Bridge of Truth” between what is claimed and what is actually happening in the rivers.

Satya-Setu is further strengthened by two sibling systems:

• Nadi-Drishti eye in the sky: a satellite-based system for river health.
• Jan-Prahari eye on the ground: a WhatsApp-powered citizen sentinel for on-ground evidence

together they provide independent satellite and crowd-sourced truth streams that Satya-Setu uses to cross-verify government data and expose “Truth Gaps.”

---

Technical Specifications

• Tech Stack:

  • Scraping & Ingestion: Python (Selenium or Playwright for dynamic dashboards; Requests + BeautifulSoup for static HTML; pdfplumber/Tesseract OCR for PDF reports).
  • Data Processing & Storage: Pandas, NumPy; flat files (CSV/Parquet) or lightweight SQLite/PostgreSQL for structured storage.
  • Backend API (Optional but recommended): FastAPI or Flask to serve unified JSON APIs.
  • Visualization: PowerBI (Public) embeds, or a custom frontend using Next.js + Chart.js / ECharts; alternatively Streamlit for faster MVP.
  • Automation & Scheduling: GitHub Actions (cron-based workflows), or low-cost alternatives like Railway/Render scheduled jobs.
  • Integration with Other Systems: Webhooks / REST API for downstream tools (e.g., Nadi-Drishti, alerting bots, newsroom dashboards).

• Core Resources & Data Sources (India-focused):

  • CPCB RTWQMS: Central Pollution Control Board’s Real-Time Water Quality Monitoring System dashboard for river monitoring stations.
  • CPCB / SPCB Portals:
    • CPCB central site (dashboards, PDF reports, station-wise data).
    • State Pollution Control Boards (SPCBs) for Ganga basin states: UPPCB (Uttar Pradesh), UKPCB (Uttarakhand), BSPCB (Bihar), WBPCB (West Bengal), etc.
    • Other key SPCBs (Karnataka, Tamil Nadu, Chhattisgarh, Maharashtra, etc.) for broader river coverage.
  • Namami Gange / NMCG Platforms:
    • PRAYAG (Platform for Real-time Analysis of Yamuna, Ganga and their Tributaries) – real-time & historical water quality and STP performance dashboards.
    • Namami Gange official project and STP progress pages.
  • India-WRIS / NWIC:
    • Water Data from National Water Informatics Centre (discharge, flow, and water level at key locations).
  • data.gov.in:
    • Datasets such as “Water Quality Data for River Ganga” and similar river datasets.
  • Other Relevant Sources (extensible list):
    • CPCB annual reports and river action plan PDFs.
    • State Environment Department reports (often in PDF/Excel).
    • Jal Jeevan Mission & Jal Shakti ministry dashboards where relevant.

---

AI Usage

• Unsupervised Anomaly Detection:

  • Implement Isolation Forests, One-Class SVMs, or Local Outlier Factor models on multivariate time-series (pH, BOD, COD, DO, turbidity, coliform counts, flow).
  • Learn the “normal” behavioral profile of each monitoring station (diurnal/seasonal variation), then flag outliers that deviate strongly from historical patterns.

• Cross-Modal Validation (Sensor Fusion Mindset):

  • Combine government sensor data with:
    • Satellite-derived indices from Nadi-Drishti (NDTI, NDCI, NDWI, turbidity proxies).
    • Rainfall and runoff data (IMD / NASA GPM) to understand when spikes are hydrologically plausible.
  • If a station reports “excellent” quality (e.g., DO high, BOD low) while satellite data shows dense algal bloom or heavy sediment plume, classify this as a “Truth Gap” anomaly.

• Meta-Data Quality Checks:

  • Detect suspicious patterns such as:
    • Repeated constant values for long periods (e.g., pH = 7.00 for 30 consecutive days).
    • Sensors that go offline at the same time every night or weekend.
    • Sudden resets to “perfect” values after being critical.

• Explainable Alerts:

  • Attach human-readable explanations to each anomaly:
    • “Unusual BOD spike compared to past 90 days and similar flow conditions.”
    • “Satellite imagery indicates turbidity increase not reflected in CPCB station data.”

---

Operation Details

• Time Estimate (MVP): ~2 Weeks (solo developer, focused scope).

  • Week 1: Scraper development for 3–5 priority sources (CPCB RTWQMS, one SPCB, PRAYAG, one data.gov.in dataset).
  • Week 2: Data model design, basic anomaly detection scripts, and a minimal dashboard (Streamlit or static site with embedded charts).

• Cost to Develop & Run (MVP):

  • Hosting: GitHub Pages (static frontend) – free.
  • Automation: GitHub Actions (limited free minutes per month) – free at small scale.
  • Compute: Local machine or a free-tier cloud instance for periodic scrapes.
  • External APIs: Using only open/public government portals and open satellite data – free.

• Monetization Strategy:

  • Premium Aggregated Data API:
    • Offer a cleaned, unified, well-documented REST API for:
      • Environmental NGOs.
      • Researchers and universities.
      • Newsrooms and data journalists.
  • Compliance & Early-Warning Alerts for Industry:
    • SaaS-style subscription where industrial clusters or individual factories get:
      • Trend alerts when their downstream water quality indicators approach legal thresholds.
      • Comparative benchmarks vs. other units in the same sector/river segment.
  • Custom Reporting:
    • Paid, on-demand analyses for PIL lawyers, policy think-tanks, or investigative journalists.

---

Detailed description

System Overview

Satya-Setu (The Bridge of Truth) is a data aggregation and anomaly detection platform focused on river and water-quality governance. It is designed for an environment where:

• Data is fragmented across dozens of portals (CPCB, SPCBs, NMCG, WRIS, PDFs, and scattered Excel sheets).
• Data is sometimes incomplete, delayed, or misreported due to faulty sensors, poor maintenance, or deliberate manipulation.
• Citizens, activists, and honest officials struggle to see the “full picture” of river health.

Satya-Setu ingests all of this into one unified system, then layers AI-driven anomaly detection and cross-validation with satellite/independent data. The goal is not only to visualize the data but to expose inconsistencies and structural blind spots.

Where Nadi-Drishti is the “eye in the sky” watching the physical river, Satya-Setu is the forensic accountant checking the books.

---

How It Works (Data & Logic Flow)

1. Data Ingestion Layer (Extract):

   • Scheduled jobs (via GitHub Actions) run at fixed intervals (e.g., hourly or daily).
   • For each source:
     • Navigate dashboards (CPCB RTWQMS, PRAYAG) with Selenium/Playwright.
     • Download CSV/Excel where available; otherwise, scrape HTML tables.
     • For PDF-based reports, download and parse them with pdfplumber + OCR.

2. Normalization & Harmonization (Transform):

   • Convert all measurements into consistent units and schemas:
     • pH (unitless, but rounding/precision standardized).
     • BOD, COD, DO (mg/L).
     • Flow/discharge (m³/s or MLD, standardized per river/segment).
   • Map diverse station identifiers into a unified station model:
     • Each station gets a canonical ID, coordinates, river name, and jurisdiction (State/District, SPCB).
   • Handle missing values and outliers:
     • Use rolling medians and interpolation for mild gaps.
     • Tag impossible values (e.g., negative BOD) as “invalid” for anomaly logs.

3. Storage & Versioning (Load):

   • Store daily snapshots (or even hourly, where available) in versioned CSV/Parquet files.
   • Maintain a simple metadata log of:
     • When data was last successfully scraped from each source.
     • How many stations responded vs. expected.

4. Anomaly Detection Layer:

   • For each station and parameter, build a historical time-series (e.g., last 365 days).
   • Train unsupervised models (Isolation Forest, LOF, or basic z-score filters) to detect:
     • Sudden jumps inconsistent with seasonal patterns.
     • Long periods of suspicious constancy.
     • Correlation breaks (e.g., BOD increasing while DO remains unrealistically high).
   • Integrate satellite-derived indicators from Nadi-Drishti for cross-validation:
     • Spatially align satellite pixels with station coordinates.
     • Create a rule base: “If satellite turbidity index is in the 95th percentile but turbidity sensor shows normal, flag a mismatch.”

5. Dashboard & Alerts:

   • The front-end presents:
     • Map View: All monitoring stations with status colors (normal / anomalous / offline / inconsistent-with-satellite).
     • Timeline View: Station-wise trend charts with anomaly points highlighted.
     • Discrepancy View: A list of places and dates where official data and satellite/derived data diverge.
   • Alerts can be exposed via:
     • Email digests.
     • Webhook callbacks for other systems.
     • Public RSS/JSON feeds.

---

Implementation Roadmap

Phase 1: Source Mapping & Prototype Scrapers (Days 1–3)

• Identify and document priority sources (for an MVP):
  • CPCB RTWQMS main dashboard.
  • One high-impact SPCB portal (e.g., Uttar Pradesh for Ganga segments).
  • PRAYAG (Namami Gange real-time portal).
  • One historical dataset from data.gov.in.
• Implement basic scrapers:
  • Target 5–10 representative stations across different states.
  • Save raw output to /data/raw/{source}/{date}.csv.

Phase 2: Data Model & Cleaning (Days 4–6)

• Design a unified schema (e.g., station_id, river, lat, lon, parameter, value, unit, timestamp, source).
• Write transformation scripts:
  • Unit normalization.
  • Handling of missing and obviously invalid values.
  • Mapping station names to consistent IDs and coordinates.
• Generate first cleaned dataset snapshot.

Phase 3: Anomaly Engine (Days 7–9)

• Implement baseline anomaly detection:
  • Simple statistical rules (e.g., z-scores, rolling mean/standard deviation).
  • Upgrade to Isolation Forest or similar unsupervised methods.
• Integrate a small subset of Nadi-Drishti satellite outputs (e.g., turbidity index at each station location).
• Define rule-based “mismatch” alerts:
  • Data present vs. satellite contradiction.
  • Data missing vs. expected reporting frequency.

Phase 4: MVP Dashboard & Public Output (Days 10–14)

• Build a thin front-end:
  • Option A: Streamlit app that reads cleaned CSVs directly from GitHub.
  • Option B: Static site (Next.js) that consumes a minimal FastAPI backend.
• Implement views:
  • River-wise and state-wise filtering.
  • Station details page with time-series charts and anomaly annotations.
  • Export options (CSV/PNG/PDF).

---

Problems Solved

A. The “Fragmented Data” Problem

• Issue: River and pollution data is scattered, inconsistent, and often difficult to access programmatically.
• Satya-Setu Solution: A single pane of glass that aggregates, cleans, and standardizes this data, making it easily queryable and comparable.

B. The “Broken or Manipulated Sensor” Problem

• Issue: Sensors can be offline, mis-calibrated, or in some cases manipulated (e.g., shutting down when discharging effluents).
• Satya-Setu Solution:
  • Tracks uptime and reporting patterns to detect suspicious behavior (e.g., nightly shutdowns).
  • Cross-validates with satellite and hydrological data to detect improbable claims (“river looks clean” vs. “sensor says dirty” or vice versa).

C. The “PDF Graveyard” Problem

• Issue: Historical data and compliance reports are usually stuck in scanned PDFs.
• Satya-Setu Solution:
  • Uses OCR and table extraction to convert legacy documents into structured datasets.
  • Makes long-term trend analysis and pre/post-intervention comparisons possible.

D. The “Lack of Accountability” Problem

• Issue: Without unified, auditable timelines of data, it is hard to prove systemic negligence or chronic pollution.
• Satya-Setu Solution:
  • Maintains versioned time-series and anomaly logs that can be cited in media investigations or legal filings.
  • Highlights specific dates, locations, and parameters where data behavior deviates from expected norms.

---

Technology Stack & Resource Requirements

Component	Technology / Tool	Cost
Scraping	Python (Selenium/Playwright, Requests, BeautifulSoup)	Open Source
Data Processing	Pandas, NumPy	Open Source
Anomaly Detection	Scikit-Learn (Isolation Forest, LOF, etc.)	Open Source
Database	CSV/Parquet on Git, or SQLite/PostgreSQL	Free/Open Src
Backend API	FastAPI / Flask	Open Source
Frontend UI	Streamlit, or Next.js + Chart.js/ECharts	Open Source
Automation	GitHub Actions	Free Tier
Hosting	GitHub Pages, Streamlit Cloud, or similar	Free Tier

---

How Technical Experts Add Unique Value

This project requires more than just web developers; it needs specific high-level expertise to handle the complexity of environmental data.

1. The Computer Vision Expert (The “Eyes”)

A specialist in Computer Vision (CV) brings the ability to “see” pollution that numerical sensors miss. Their role is to turn satellite imagery into actionable data points.

• Sensor Fusion (The “Trust but Verify” Logic):
  • Concept: A single data source is rarely perfect. A government sensor might report “Clean Water” because it is broken or placed in a clean pocket of the river.
  • Application: The CV expert builds a Multi-Modal Validation System. They fuse the numerical data (from ground sensors) with visual data (satellite imagery). If the sensor says “Low Turbidity” but the satellite sees a massive brown plume, the system flags a “Truth Gap” anomaly.
• Pattern Recognition Beyond Human Scale:
  • Concept: Pollution has visual “fingerprints”—oil slicks look different from algal blooms, which look different from mud.
  • Application: They train Deep Learning models (like U-Nets or CNNs) to automatically classify these textures. The system learns to distinguish between a natural sediment flow (after rain) and an industrial chemical discharge (which might be an unnatural color like dark black or foam-white), even in areas where no ground sensors exist.

2. The Data Scraping & Engineering Expert (The “Memory”)

A specialist in advanced web scraping and data engineering builds the pipeline that keeps the system alive. Their role is to ensure we have a continuous, clean stream of data from hundreds of fragmented sources.

• Resilience Against “Data Rot”:
  • Concept: Government portals are notoriously unstable. They change layouts, go offline, or hide data behind captchas.
  • Application: The expert builds Self-Healing Scrapers. If the CPCB website changes its table structure, the scraper detects the break and attempts to parse the new layout automatically, or alerts the team instantly. They use Headless Browsers (like Playwright/Selenium) to navigate complex, dynamic dashboards that standard scripts can’t read.
• Unsupervised Anomaly Detection:
  • Concept: We don’t always know what a “pollution event” looks like mathematically, but we know what “normal” looks like.
  • Application: The expert implements algorithms (like Isolation Forests) that learn the “heartbeat” of every monitoring station. If a factory suddenly releases a massive load of effluent at 3 AM, the data pattern will deviate from the historical norm. The system flags this “statistical noise” as a probable illegal discharge, even if the specific pollutant wasn’t explicitly programmed into the rules.

---

Beneficiary Experts (Target Audience)

This tool is designed to assist specific categories of experts and organizations, with a primary focus on the Mahanadi and Kaveri basins.

In India (Mahanadi & Kaveri Focus)

• State & Regional Regulators:
  • Odisha & Chhattisgarh SPCB Officials: Who need to monitor the heavy industrial belt (Iron, Steel, Power) in the Mahanadi basin without relying solely on factory self-reports.
  • Tamil Nadu & Karnataka PCB Officials: Who need real-time alerts for textile dyeing effluent in the Noyyal/Kaveri belt (Erode/Tiruppur) and sugar distillery waste in Mandya.
• River Basin Authorities:
  • Mahanadi River Basin Organization & cMahanadi (NIT Raipur/Rourkela): Researchers needing integrated data to plan the “Mahanadi River Basin Management Plan.”
  • Cauvery Water Management Authority (CWMA): Officials monitoring flow and quality disputes between states.
• Environmental Journalists:
  • Reporters in Cuttack, Sambalpur, or Erode who need hard data to expose illegal dumping during monsoon seasons.
• Public Interest Litigators:
  • Legal teams fighting cases in the NGT (National Green Tribunal) regarding fly ash disposal in Talcher or untreated sewage in Trichy.

Around the World (Global Context)

• International Water Security Researchers: Academics studying the decline of major river deltas due to climate change.
• ESG (Environmental, Social, and Governance) Auditors: Firms that audit supply chains for multinational corporations, verifying that their factories are not polluting local water bodies.
• Conservation NGOs: Organizations like the WWF or The Nature Conservancy that require data to lobby for policy changes without the budget for private satellite contracts.
• Climate Change Data Scientists: Researchers feeding real-world surface water data into global climate models.

---

Market analysis and Business plan

Competitive Analysis

The market for environmental data is crowded, but heavily segmented. Most players focus on Air Quality (because it affects rich urbanites) or Global Climate Trends (for insurance/finance). Very few focus on hyper-local river compliance in India’s industrial hinterlands.

1. The “Open Data” Aggregators (Civic Tech)

These are our closest functional relatives—platforms that scrape government data for public good.

OpenAQ (Global):

• What they do: The gold standard for air quality. They scrape government portals worldwide (including CPCB) to create a single, harmonized API for PM2.5/PM10.
• Weakness:
  • Air-Only: They do not touch Water Quality. The complexities of water (river flow vs. BOD vs. pH vs. heavy metals) are much harder to standardize than simple air particulates.
  • No “Actionable” Intelligence: They provide the data “as is.” They don’t have a “Lie Detector” layer that says, “This sensor is likely broken because the value has been constant for 48 hours.”

OpenCity.in / Oorvani Foundation (India):

• What they do: An excellent repository of urban data (Bengaluru/Chennai), including some lake water quality reports (PDFs from KSPCB).****
• Weakness:
  • Static & Manual: They mostly host static files (PDFs/CSVs) uploaded manually or semi-manually. They don’t have real-time, minute-by-minute scrapers running against industrial effluent dashboards.
  • Urban Bias: Focus is on cities (Bangalore Lakes), not major river basins like the Mahanadi or Kaveri industrial belts

• What they do: They build “Command & Control” centers for state governments (e.g., Andhra Pradesh Water Resources Dept). They integrate IoT sensors and hydrological models for flood management. [web:78][web:84]
• Weakness: They are a B2G (Business-to-Government) partner. This means they are part of the system. They cannot serve as a “Watchdog” because the government is their client. If a state-owned thermal plant is polluting, Vassar’s dashboard is internal to the government and unlikely to flag it publicly.
• Satya-Setu Advantage: Independence. We are an external auditor. We serve the public, not the polluter.

India Water Portal / Arghyam:

• What they do: A massive knowledge repository for water. They host articles and some datasets.****
• Weakness: It is a “Library,” not a “Live Dashboard.” You go there to read a report from 2023, not to see what NALCO discharged into the Mahanadi 10 minutes ago.

2. The Government Portals (Our Raw Material)

CPCB Central Portal & Ganga PRAYAG:

• What they do: The official repository of Real-Time Water Quality Monitoring Systems (RTWQMS). It aggregates data from ~4,000 industrial sensors.
• Weakness:
  • The “Friday Night Glitch”: Sensors notoriously go “offline” or show flat-line data during illegal dumping hours. The portal accepts this “No Data” status passively.
  • Fragmentation: Data for the Mahanadi (Odisha/Chhattisgarh) is often on a completely different URL than data for the Kaveri (TN/Karnataka). There is no single API to query both.
• Satya-Setu Advantage: Active Anomaly Detection. We don’t just display “No Data.” We flag “No Data” as a suspicious event if it correlates with satellite-detected plumes or occurs repeatedly at specific times.

3. Global & Academic Databases

WRI Aqueduct (World Resources Institute):

• What they do: A global water risk atlas used by investors to see if a region has high water stress.
• Weakness: It relies on annual or monthly averages. It cannot tell you that “Factory X dumped sludge yesterday at 3 AM.” It is a strategic planning tool, not a tactical policing tool.
• Satya-Setu Advantage: Real-time tactical intelligence. We operate on a 24-hour cycle, not an annual cycle.

4. The “Compliance Intelligence” Firms (Private/Dark Data)

These companies scrape data to sell “Risk Signals” to banks and investors.

Respirer Living Sciences (India):

• What they do: They started with low-cost air sensors but now aggregate CPCB data for their “NCAP Tracker” and “Atlas AQ” dashboard. They have successfully built a business on top of government data.
• Weakness:
  • Hardware Heavy: A lot of their business is still selling their own sensors (Atmos). Satya-Setu avoids the hardware logistics trap entirely.
  • Air Focus: Like OpenAQ, their mature products are Air-centric. Water is a secondary frontier they haven’t fully conquered with the same depth.

Generic “Alternative Data” Scrapers (Bright Data / Import.io users):

• What they do: Hedge funds hire teams to scrape specific factories to predict production output (e.g., “If the factory is using more water, they are producing more steel”).
• Weakness:
  • Opaque & Expensive: This data never sees the light of day. It is sold privately to traders. Satya-Setu democratizes this “Insider Information” for the public.

How Satya-Setu is Better (The “Lie Detector” Moat)

Most scrapers (like OpenAQ) assume the government data is true . They just make it readable . Satya-Setu assumes the data might be false.

Feature	OpenAQ / OpenCity	Government Portals	Satya-Setu
Data Trust	”Trust the Source"	"We are the Source"	"Trust but Verify” (Zero Trust)
Handling PDFs	Host them as files	Publish them	OCR & Digitize them into Graphs
Context	Raw Numbers	Raw Numbers	Context (e.g., “This pH is illegal”)
Cross-Check	None	None	Satellite vs. Sensor Comparison
Cost	Free (Grant funded)	Public	Free (Public) + Paid API (Business)

The Business Plan (Zero to Scale)

Phase 1: The “Truth Prototype” (Months 1–3)

• Goal: Build a “Watchdog Dashboard” for 2 critical stretches:
  1. Mahanadi: The industrial corridor from Sambalpur to Cuttack (monitoring NALCO, NTPC, and Coal belts).
  2. Kaveri: The textile belt from Erode to Karur.
• Data Sources: Scrape data from OSPCB (Odisha), CECB (Chhattisgarh), and TNPCB (Tamil Nadu) portals.
• Customer: Free for local activists and cMahanadi researchers to build credibility.

Phase 2: Corporate Accountability & CSR (Months 4–9)

• Goal: Monetize via “Impact Validation” reports.
• Pitch to Industry: “You spend crores on CSR. Use our dashboard to prove to shareholders that your initiatives are actually cleaning the river.”
• Target Revenue: ₹5 Lakh - ₹15 Lakh per customized river-audit report.

Phase 3: The “Green Passport” (Supply Chain Assurance) (Year 1+)

Replaces the “Compliance Shield” with a “Verification Engine.”

Instead of helping factories avoid fines (which is low-impact), we help them win business by proving they are clean. We target the Global Brands and Banks that hold the power.

• Goal: Create a “Satya-Setu Verified” data layer for the Global Supply Chain.
• The Shift: We don’t ask the factory for money to hide data. We ask the Buyer (e.g., H&M, Tata Motors) for money to verify the factory’s data.

The “Green Passport” Product Line

1. For Global Buyers (The “Tiruppur Model”) Target: Fashion Brands (H&M, Zara, Decathlon) sourcing from the Kaveri Basin.

• Problem: These brands have pledged “Zero Discharge” (ZDHC) goals. They currently rely on annual paper audits, which are easy to fake.
• Our Solution: A “Live Water Audit” API.
  • When a brand places an order for 10,000 shirts, they query Satya-Setu: “Did Supplier Factory #402 have any illegal discharge alerts during the production month?”
  • If Satya-Setu detects a “Truth Gap” (Satellite showed plume vs. Sensor showed clean), the brand gets a Risk Alert.
  • Impact: The factory must clean up to keep the contract.

2. For Green Finance (The “Mahanadi Model”) Target: Banks (SBI, Axis) & Green Bond Issuers lending to Steel/Power plants.

• Problem: Banks are issuing “Sustainability-Linked Loans” (SLLs) where interest rates drop if the factory meets environmental targets. Currently, banks lack data to verify if targets are met.
• Our Solution: Dynamic Credit Scoring.
  • We provide a “Water Risk Score” for every borrower in the Mahanadi belt.
  • If a steel plant consistently accurately reports data (Low Anomaly Score), their risk rating improves.
  • Impact: Pollution becomes a financial liability. Clean factories get cheaper loans.

3. The “Satya-Setu Certificate” (QR Code)

• A public-facing QR code for factories to display on their website.
• Scans show a “Real-Time Trust Score” based on their last 30 days of data transparency (Uptime + Consistency).
• Monetization: Factories pay a certification fee to display this badge, validated by our independent scraping.

Why This Business Model Wins

1. Cleaner Incentives: We align profit with cleanliness, not secrecy.
2. Deeper Pockets: Global brands and banks have millions in “Sustainability Budgets,” whereas local factories have tight margins.
3. Adversarial Truth: We don’t need the factory’s permission. We audit them from the outside (Satellite + Public Data) and sell the truth to their financiers.

---

Possible Customers (Mahanadi & Kaveri Specific)

These are the specific entities we will target for Phase 2 (CSR Reports) and Phase 3 (Compliance Subscriptions).

A. Mahanadi Basin (Odisha & Chhattisgarh)

Focus: Heavy Metal, Thermal Pollution, Fly Ash

1. Mining & Metal Giants (The “Big Fish”):

   • NALCO (National Aluminium Company): Angul & Damanjodi units.
   • SAIL - Bhilai Steel Plant: Major water consumer and potential effluent source in Chhattisgarh.
   • Vedanta Aluminium: Jharsuguda & Lanjigarh refineries.
   • Jindal Steel & Power Ltd (JSPL): Angul.
   • Hindalco Industries: Hirakud.

2. Power Generation (Thermal Plumes):

   • NTPC Talcher & NTPC Kaniha: Large thermal power plants that discharge heated water (thermal pollution).
   • Mahanadi Coalfields Limited (MCL): Opencast mines affecting groundwater and surface runoff.

3. Fertilizers:

   • IFFCO Paradip: Phosphatic fertilizer plant near the delta.

B. Kaveri Basin (Karnataka & Tamil Nadu)

Focus: Chemical Dyeing, Organic Waste, Salinity

1. Textile & Paper (The “Chemical” Cluster):

   • Tamil Nadu Newsprint and Papers Limited (TNPL): Karur (Large water consumer).
   • Seshasayee Paper and Boards Ltd: Erode.
   • Tiruppur Exporters’ Association: Represents hundreds of dyeing units (potential bulk subscription for the association).

2. Sugar & Distilleries (The “Organic” Load):

   • E.I.D. Parry (India) Ltd: Sugar factories in the basin.
   • Mysore Sugar Company: Mandya, Karnataka.
   • Bannari Amman Sugars: Operations in Tamil Nadu.

3. Cement & Chemicals:

   • Chemplast Sanmar: Mettur (PVC/Chemicals).
   • Chettinad Cement / Dalmia Cement: Ariyalur belt.

C. Institutional & Government Partners

For Grants, Data Partnerships, and Pilot Projects

1. cMahanadi (Center for Mahanadi River Basin Management): A consortium of NIT Raipur & NIT Rourkela; ideal for academic collaboration and credible data endorsement
2. Cauvery Calling (Isha Foundation) / Art of Living: Large NGOs with massive river rejuvenation funds that need “Proof of Work” data.
3. Odisha State Pollution Control Board (OSPCB): Direct collaboration for the “Star Rating” program for industries.