FSCL Analytics – Data Analytics Platform for Smart Water Distribution Systems

Overview

The FSCL Analytics project is a large-scale data analytics platform designed to support smart city infrastructure with a primary focus on water distribution monitoring and analysis. In modern urban planning, smart water management systems are key to ensuring sustainability, efficiency, and equitable resource distribution. These systems, powered by automation technologies like PLC (Programmable Logic Controller), SCADA (Supervisory Control and Data Acquisition), and IoT-enabled sensors, generate massive volumes of data continuously. However, collecting this data is only the beginning; the real value lies in how this data is processed, analyzed, and interpreted for actionable decision-making.

In this specific implementation, FSCL Analytics serves as a centralized analytics solution to collect, store, and visualize data from more than 90 water distribution sites. These sites feed in time-series data at intervals of 5 to 30 minutes from thousands of sensors measuring water flow, pipeline pressure, motor status, and energy consumption. Over the course of a month, this leads to over 3.45 crore rows of data, and over a five-year operational window, the platform is expected to handle more than 200 crore rows.

The overarching goal of the FSCL Analytics platform is to provide city authorities, municipal corporations, and operational managers with a reliable and insightful system to monitor resource usage, detect inefficiencies, and support long-term planning and optimization efforts. The project combines real-time monitoring with historical data analytics, offering site-wise, month-wise, and year-wise comparative insights into water usage trends.

Problem Statement

Urban water management is one of the most critical challenges faced by municipal bodies today, especially in densely populated cities. As cities grow, so does the complexity of their water infrastructure. While deploying PLCs and SCADA systems across the water network helps automate the operations, they also generate vast volumes of telemetry and control data that can be overwhelming without the right data processing infrastructure.

The problem faced in FSCL’s scenario is two-fold:

Data Accumulation and Storage:
- The system gathers over 11.4 lakh rows of data per day.
- This translates to 3.45 crore rows in a month and over 200 crore rows in five years.
- Managing, storing, and securing this volume of time-series data requires high-performance hardware and scalable storage solutions.
Real-time and Historical Analysis:
- Authorities require month-wise comparison of water flow, energy consumption, and pump utilization to identify usage patterns, seasonal trends, and anomalies.
- Performing even simple queries on recent data can take up to 20 minutes, even with a professional-grade server setup (40-core CPU, 128 GB RAM, SATA HDD with 12 GBPS bandwidth).
- This sluggish performance hinders the ability to make timely decisions, especially in scenarios requiring quick response like leak detection, abnormal consumption spikes, or pump failure alerts.

Thus, the challenge is not merely about capturing data but about making it usable and accessible at scale — a classic data engineering and analytics problem.

Description

The FSCL Analytics project is built around the real-world scenario of managing a city’s water distribution system through automation and intelligent data analytics. Here’s a deeper breakdown of how the system is designed and what it encompasses:

Categorization of Sites

The 90+ water distribution sites in the city are classified into the following four categories based on their function in the water supply chain:

Category	Description
Renny wells	Primary sources where raw water is received and stored before distribution.
Main Boosting Stations (MBS)	Major hubs responsible for pumping water to large city zones.
Intermediate Boosting Stations (IBS)	Mid-level stations that receive water from MBS and distribute it further.
Sub Boosting Stations (SBS)	Smaller, localized stations that ensure water pressure and delivery at the final mile.

This classification helps streamline data monitoring, maintenance scheduling, and performance analysis specific to each level in the distribution hierarchy.

Site-Level Data Capture

Each of the 90+ sites includes:

Inflow Monitoring: Water is channeled into underground tanks with an average storage capacity of 40 lakh liters (4 million liters). The inflow is constantly monitored using flow meters to ensure that adequate water is being received.
Header-Based Distribution: Water is distributed via 2 to 3 pipeline headers per site. Each header has a pipe diameter of 200 to 300 mm and is fitted with flow meters to monitor output. These flow meters provide real-time data about how much water is being delivered to different zones.
Pressure Sensors: Pressure data is crucial for identifying blockages, leakages, or pump inefficiencies. All pressure readings are captured continuously and logged.
Motor and Pump Monitoring: Motors connected to the storage tanks are responsible for distributing the water. Local PLCs monitor each motor’s run hours, operational cycles, and response to control signals.
Energy Metering: Each site also includes energy meters that connect to PLCs via Modbus protocols. These meters track voltage, current, and power consumed by the pumps — valuable data for both energy efficiency tracking and predictive maintenance.

Asset Distribution Table

The FSCL Analytics platform encompasses a wide range of field assets deployed across all water distribution sites to ensure comprehensive monitoring and control. These assets include flowmeters for tracking water transmission, level transmitters for monitoring tank levels, and energy meters for capturing power usage data. Additionally, motors are installed to manage water pumping operations, while pressure transmitters help in identifying pipeline anomalies such as leaks or blockages. This robust asset infrastructure forms the backbone of accurate data acquisition and smart decision-making across the system.

Asset Name	Total Count
Flowmeter	234
Pumps/Motors	531
Level Transmitters (LT)	96
Pressure Transmitter (PT)	684
Energy Meter	95

Data Collection and Transmission

Data from each sensor and meter is collected at intervals ranging from 5 to 30 minutes.

Over 4000+ unique data points across the city feed into the analytics system.
Data is sent to a central server over secure protocols where it is stored in structured databases optimized for time-series analysis.

Data Volume

Let’s look at the data scale:

Daily Data Rows: ~11.4 lakh (1.14 million)
Monthly Data Rows: ~3.45 crore (34.5 million)
Five-Year Data Rows: ~200 crore (2 billion+)

This makes FSCL Analytics one of the most data-intensive municipal monitoring platforms, rivaling the scale of many enterprise-grade IoT solutions.

Infrastructure Challenges

Despite having a high-performance server setup:

CPU: 40 cores
RAM: 128 GB
Disk Bandwidth: 12 GBPS SATA HDD

Even simple analytical queries — like comparing water flow trends over the past month at a single site — can take up to 20 minutes. This is unacceptable for real-time decision-making and emergency responses.

The volume of data, while rich with potential insights, poses a significant computational challenge. The current hardware, although powerful, is not sufficient when data indexing, optimization, and efficient querying are not properly engineered.

Conclusion

The FSCL Analytics project is a compelling case study in real-world industrial data analytics at scale. It shows the transition from automation to intelligent decision-making through data.

Key Takeaways:

From Automation to Insight: Automation via SCADA and PLCs provides the infrastructure for data collection. FSCL Analytics takes the next step — transforming that data into actionable insights.
Scale of Operation: Managing data from 90+ geographically distributed sites, with over 4000 sensor points feeding real-time information, is no small feat. The system’s ability to handle this scale over a multi-year period is a testament to robust engineering.
Data Engineering Matters: The major bottleneck in the system is not just hardware but the lack of optimized data engineering — indexing strategies, compression, partitioning, and potentially distributed processing frameworks like Apache Spark or Presto could drastically improve performance.
Need for Evolution: As data volume grows, so must the intelligence of the system. Moving towards cloud-native solutions with elastic scaling, parallel processing, and in-memory analytics might be the future roadmap.

Ultimately, the FSCL Analytics project highlights the potential of leveraging data in building smarter cities, while also shedding light on the real-world limitations and challenges of scaling data systems. Whether it’s tracking water usage, optimizing energy consumption, or planning infrastructure investments, data is at the heart of every smart decision. The platform’s continued evolution will be essential in ensuring that insights are delivered at the speed of thought — or faster.

FSCL Analytics Project Technical Summary

FSCL Raw database “HistData” is transformed to useful data to analyze the daily/monthly/yearly water in and water out data points into FSCL Analytics database. Used

The below technical Skills are involved in this project

.Net Angular- Web and API
Microsoft SQL Server 20222 Standard Edition

Created SQL Server stored procedures scripts for each lines to perform the ETL (Extract Transform Load) to load raw database table records intodaily useful analytics records in the form of daily water in and out in ML.Find the below screen shot for SQL Server stored procedure details.

Also, created a SQL Server jobs to execute every 20mins to sync the data from HistData database to FSCLAnalytics database tables. Find the below screen shot for SQL Server jobs details.