Smart Water Distribution and Management System using IoT in

Smart Cities

A report on Project Phase – I submitted

in partial fulfilment of the requirements for the award of the degree of

Bachelor of Technology
in
Computer Science and Engineering
by
GANTA VARA PRASAD
Registration No.: 2020105238

SUSHMA R
Registration No.: 2020105256

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

NATIONAL INSTITUTE OF TECHNOLOGY NAGALAND

DECEMBER 2023
 i

Smart Water Distribution and Management System using IoT in

Smart Cities

A report on Project Phase – I submitted

in partial fulfilment of the requirements for the award of the degree of

Bachelor of Technology
in
Computer Science and Engineering
by
GANTA VARA PRASAD
Registration No.: 2020105238

SUSHMA R
Registration No.: 2020105256

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

NATIONAL INSTITUTE OF TECHNOLOGY NAGALAND

DECEMBER 2023
 ii

रा ीय ौ ोिगकी सं ान नागालड
NATIONAL INSTITUTE OF TECHNOLOGY NAGALAND
DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING
Chumukedima, Dimapur – 797103, Nagaland

BONAFIDE CERTIFICATE

Certified that this Project titled “Smart Water Distribution and

Management System using IoT in Smart Cities” is the bonafide work of

Ganta Vara Prasad (2020105238) & Sushma R (2020105256) who carried out the

work under my supervision. Certified further, that to the best of my knowledge the

work reported herein does not form part of any other project or dissertation on the

basis of which a degree or award was conferred on an earlier occasion on this or any

other student.

Dr. ARAMBAM NEELIMA Dr. NAGARAJU BAYDETI

Project Coordinator Project Supervisor
Assistant Professor Assistant Professor
Department of Computer Department of Computer
Science and Engineering Science and Engineering
NIT Nagaland NIT Nagaland

Dr. LITHUNGO MURRY Dr. H. ROHEN SINGH

Assistant Professor External Examiner
Department of Computer Assistant Professor
Science and Engineering Department of IT School of
NIT Nagaland Engineering and Technology,
Nagaland University
 iii

ABSTRACT

Water is absolutely necessary for our daily existence. With the growing
demand, we also see that the availability of water is diminishing. A significant portion
of the global population lacks access to fresh water. Numerous factors, including
extreme weather conditions, excessive usage, complete urbanization, industrial
expansion, lack of proper infrastructure, etc., are blamed for water scarcity. In housing
societies, challenges arise from the uniform distribution of water irrespective of the
consumption rate, leaks during transportation, and the energy-intensive supply of
water from the main source. In response to these challenges, we have adopted Internet
of Things (IoT) technology for monitoring water levels, detecting leaks, and
implementing automated tank refilling when required. We initiate the process by
calculating the daily consumption rate and distributing water accordingly to pre-
emptively prevent water scarcity. In the event of any shortage, rather than supplying
water from the main source, we redirect the supply from the nearest water tank whose
consumption rate is lower than the daily consumption rate. This approach enhances
water resource management and reduces the dependency on the primary water source.
 iv

ACKNOWLEDGEMENT

We wish to place on record our deep sense of gratitude to our honorific Guide
Dr. Nagaraju Baydeti, Assistant Professor, National Institute of Technology
Nagaland for his supervision, valuable guidance and moral support leading to the
successful completion of the work. Without his continuous encouragement and
involvement, this project would not have been a reality.

We would like to extend our sincere thanks to Dr. Arambam Neelima for
their valuable suggestions. We wish to thank Dr. Lithungo Murry, Head of the
Department Computer Science and Engineering, NIT Nagaland for continuous support.
We would also like to thank all our friends who have developed us to gain a sense of
dutifulness, perfection and sincerity in the effort.

We wish to dedicate this work to parents and teachers, for they are the
pillars of support giving us confidence in whatever we do. We would like to thank
Saurabh Kumar, PhD Scholar and Shyam Sunder Jannu Soloman, PhD Scholar who
has motivated us to work harder and do our best. Last but not least, we would like to
owe our sincere and incessant gratitude to the almighty God for the immense blessing
on us.

GANTA VARA PRASAD

(2020105238)

&

SUSHMA R

(2020105256)
 v

TABLE OF CONTENTS

CHAPTER NO. TITLE PAGE NO.

ABSTRACT iii
LIST OF FIGURES vii
LIST OF ABBREVIATIONS viii

1. INTRODUCTION 1-6
1.1 GENERAL 1
1.2 STATE OF ART 2
1.3 OBJECTIVES 5
1.4 ORGANIZATION OF THE REPORT 6

2. DATA COLLECTION AND EXPLORATORY ANALYSIS 7-9

2.1 APPROACH AND METHODOLOGY 7
2.2 DATASET 8
2.3 EXPLORATORY DATA ANALYSIS 9

3. PROPOSED MODEL 10-17

3.1 INTRODUCTION 10
3.2 MODULES 12-17
3.2.1 WATER USAGE MANAGEMENT 12
3.2.2 LEAKAGE DETECTION 13
3.2.3 SHORTEST PATH 15
3.2.4 CONTROLLING VALVES 16
 vi

CHAPTER NO. TITLE PAGE NO.

4. RESULTS AND DISCUSSION 18-20

4.1 INTRODUCTION 18
4.2 PERFORMANCE 19

5. CONCLUSION 21-22
5.1 INTRODUCTION 21
5.2 HIGHLIGHTS OF THE WORK 21
5.3 FUTURE WORK 22
 vii

LIST OF FIGURES

FIGURE NO. TITLE PAGE

NO.

3.1 SMART WATER DISTRIBUTION NETWORK 9

3.2 WATER DEMAND FORECASTING 10
3.3 WATER MANAGEMENT 10
4.1 HOURLY WATER CONSUMPTION OF TANK 1 18
4.2 HOURLY WATER CONSUMPTION OF TANK 2 18
4.3 HOURLY WATER CONSUMPTION OF TANK 3 18
4.4 HOURLY WATER CONSUMPTION OF TANK 4 18
4.5 HOURLY WATER CONSUMPTION OF TANK 5 18
4.6 WATER USAGE ON 02-11-2023 18
4.7 INTERNAL WATER TRANSFERS ON 02-11-2023 19
4.8 HOURLY WATER CONSUMPTION OF TANK 1 19
4.9 HOURLY WATER CONSUMPTION OF TANK 2 19
4.10 HOURLY WATER CONSUMPTION OF TANK 3 19
4.11 HOURLY WATER CONSUMPTION OF TANK 4 19
4.12 HOURLY WATER CONSUMPTION OF TANK 5 19
4.13 WATER USAGE ON 03-12-2023 19
4.14 INTERNAL WATER TRANSFERS ON 03-12-2023 19
 viii

LIST OF ABBREVIATIONS

IoT - Internet of Things

AIRMA - AutoRegressive Integrated Moving Average
EPANET - Environmental Protection Agency Network
Evaluation Tool
MQTT - Message Queuing Telemetry Transport
LIDAR - Light Detection and Ranging
CNN - Convolutional Neural Network
GRU - Gated Recurrent Unit
MAPE - Mean Absolute Percentage Error
STLF - Short-Term Load Forecasting
BPNN - Backpropagation Neural Network
RMSE - Root Mean Square Error
MSE - Mean Squared Error
CBGRU - Convolutional Bidirectional Rated Recurrent Unit
£ - Tuning parameter
 1

CHAPTER 1

INTRODUCTION

1.1 GENERAL
Water is vitally important for the existence of all living organisms. It is
important in all aspects such as agriculture, industrial processes, and climate regulation,
and also is a great support to the ecosystem. With the growing population, there is also
an increase in water demand. Beyond its impact on vast rural and urban populations,
water scarcity in India has far-reaching consequences for the environment and
agricultural sector. Despite a population exceeding 1.4 billion people, India possesses
merely 4% of the world’s freshwater resources. India is facing increasing challenges
related to water scarcity due to various factors such as the increase in population,
changes in climatic patterns, uneven distribution of water across the country, and water
pollution which further reduces the availability of clean water. Ensuring India’s
economic and social well-being requires the availability of sufficient resources to meet
the needs of agriculture, industries, and the domestic sector in the years ahead.

In recent years, numerous occurrences have brought attention to the water challenges
in India, revealing the gravity of the issue. One of the noticeable incidents includes:

Chennai Water Crisis (2019): On June 19, 2019, authorities in Chennai officially
announced the arrival of “Day Zero,” or the day when almost no water is left.
This declaration came as all four primary reservoirs that ran throughout the city
had depleted due to two consecutive years of inadequate monsoon rainfall,
especially in late 2017 and throughout a significant portion of 2018.

The primary issue in water scarcity arises from the equitable distribution of
water without accounting for consumption. The equal distribution results in regional
tanks with higher water consumption facing shortages.
 2

Additionally, water scarcity can be attributed to losses during transportation.

Furthermore, supplying water from the main source when shortages occur proves to be
energy-intensive. Therefore, the issue is not solely about water scarcity, it also involves
the need for more smart and efficient water usage.

1.2 STATE OF THE ART

The availability of fresh water is diminishing every year. According to an

economic forum survey, increasing water demand will lead to significant scarcity
worldwide in the next two decades. It is tough to stop the growing need for water and
prevent loss while transporting it. Lakshmi Kathan Narayanan et al. [1] proposed an
IoT-based architecture integrated with fog computing. In this proposed model first
demand forecasting is carried out using Autoregressive Integrated Moving Average
(ARIMA). Based on water demand forecasting analysis, IoT-based architecture is
designed for the proper distribution of water. Fog computing is used for data collection
from IoT devices and EPANET is used for simulation. As a result, this model can
simulate underground water by combining IoT and demand forecasting, achieving a
Mean Absolute Percentage Error of 52.45%.

Ankit Anilkumar Maroli et al. [2] introduced a framework aimed at addressing

the water scarcity issues in rural India caused by inadequate infrastructure and
mismanagement of water resources. The prototype is developed by analyzing the
village and its water demand and availability of water resources. IoT devices like flow
sensors and ultrasonic level sensors are used to minimize loss. As a result, a complete
framework for efficient water storage and distribution is developed which includes
finding and fixing leaks.
 3

M. M. Srihari et al. [3] suggested a system for smart water management system
that includes features such as checking water quality and detecting water leaks. In this
model water level sensor is used to regulate water flow, a Ph sensor is used to check
the Ph value of the water, and a flow sensor is used to detect leaks if any. All these
sensors are linked directly to the Arduino board and the IoT module employed is used
to link the Arduino board to the internet, retrieves information from IoT devices, and
transfers it using MQTT protocol.

Kaushik Gupta et al. [4] proposed a smart water management system that attains
full automation. In this proposed model they used an ultrasonic sensor and turbidity
sensor for observing the level and purity of water, these data are transmitted
continuously using MMQT protocol. Two Raspberry Pi Zero W are used to interpret
data and relay switches are used to control the water from the smartphone application
thus achieving complete automation.

Thomas Robles et al. [5] suggested an updated design for smart water
management by incorporating IoT technologies to separate decision support systems
and monitoring from business processes coordination and subsystem implementation.
Here a MEGA model architecture is used which sets a standard protocol to decouple
all water management globally. In this model, interoperability between the interfaces
is provided.

Farmanoullah Jan et al. [6] presented an examination of IoT-managed water

storage tanks, concentrating on monitoring water levels, identifying leaks, and auto-
refilling water tanks, particularly at the individual level. For monitoring water levels,
commonly used devices include ultrasonic sensors, magnetic float sensors, and
LIDAR. Detecting water leakages involves placing a pair of water flow sensors at the
starting and endpoints of the pipe. The automatic refilling of tanks can be facilitated
using relays. To capture sensor data, and transfer it to a cloud server via local gateway
or Ethernet, components like Arduino Uno R3, Raspberry Pi 4, etc., are required.
 4

Henry Fuentes et al. [7] introduced a framework for a system dedicated to

measuring water consumption considering five key aspects: water consumption, local
record consumption process, physical security of the electronic device, storage, and
visualization of the consumption obtained, and leakage detection. The recorded water
consumption is transmitted to the local server Gateway via Wi-Fi, utilizing the
lightweight MQTT protocol. Upon receiving the data, processing occurs through the
“Node-Red” software. In case of any physical tampering with the “Raspberry Pi”
device, an alert is generated to ensure its proper operation using a vibration sensor such
as SW-420. The data collected from the “Edge Gateway” is stored in a NoSQL database
named “Cloudant.” The water leakage algorithm identifies potential leaks by
considering four scenarios, utilizing input parameters such as device ID, user ID, time
(t2), consumption at t2, and location.

Mingyang Pan et al. [8] employed deep neural networks to explore the
prediction of water levels. The primary emphasis of the paper is on utilizing this
extensive dataset to develop deep neural network models for accurate predictions. To
enhance prediction accuracy, a hybrid CNN-GRU model is constructed, combining the
gated recurrent unit (GRU) to capture changing water level trends and the
convolutional neural network (CNN) to discern spatial correlations among water level
data from neighbouring stations.

Lizhen Wu et al. [9] proposed a GRU-CNN hybrid neural network combining

the strengths of the GRU and CNN models as a means to enhance the accuracy of
Short-Term Load Forecasting (STLF). The GRU model focuses on processing time
sequence data, while the CNN module excels in handling spatiotemporal matrices.
When the performances of the proposed GRU-CNN model were compared with
BPNN, GRU, and CNN models, results indicated that it has the lowest values of MAPE
and RMSE thus achieving the best performance among all the four models.
 5

Muhammad Sajjad et al. [10] proposed a hybrid CNN-GRU model introduced

for forecasting electricity consumption in residential buildings. Given the non-linearity
of the input data, a standard min-max scalar normalization is applied before performing
the training process. A hybrid model is created for training which involves initially
extracting spatial features using CNN and inputting them into a multi-layered GRU.

Qing Tao et al. [11] introduced a deep learning-based short-term forecasting

model for PM2.5 concentration, presenting the CBGRU method which combines 1D
convnets and bidirectional GRU neural networks. The results reveal that the CBGRU
model achieves lower error and superior prediction performance.

Liyuan Denga et al. [12] Studied a novel multivariate time-series prediction

model utilizing a combination of CNN and GRU for accurate water consumption
prediction. The proposed model initiates with the application of CNN to extract robust
features, thereby simplifying the complexity of original data. These extracted features
are then fed to GRU neural network. This model uses MSE as target loss function
which enables the model to produce the output of fully connected layer, which is
denormalized to obtain the predicted value of next day.

1.3 OBJECTIVES

We will address the mentioned issues by employing IoT devices and Machine
Learning algorithms to calculate the daily consumption rate. Based on the predicted
data, water distribution is managed to ensure there is no water shortage. To counter the
loss due to transportation, flow sensors are used they are strategically placed at pipe
endpoints to measure flow, detecting any potential leaks. If a shortage is identified,
rather than supplying water from the main source, we can redirect it from the nearest
tank with a lower consumption rate, thereby reducing energy consumption.
 6

Specifically, the objectives of this paper include:

 To predict the usage of water and distribute it accordingly using IoT and
machine learning algorithms.
 To recognize the leakage of water from pipes while transporting using flow
sensors.
 Achieving the task of supplying water to the water tank having water underneath
the minimum water level, from the nearest water tank where its consumption is
lower as compared to its daily consumption.

1.4 ORGANIZATION OF THE PROJECT

The structure of the paper is as follows:

Chapter 1: Provides a concise overview of the project.

Chapter 2: Shows the data collection and exploratory data analysis

Chapter 3: Outlines the model proposed for the project.

Chapter 4: Presents the outcomes of the project and engages in a discussion around

them.

Chapter 5: Summarizes the key findings and outlines potential directions for future

research.
 7

CHAPTER 2

DATA COLLECTION AND EXPLORATORY ANALYSIS

2.1 APPROACH AND METHODOLOGY

PyCharm, crafted by JetBrains, is a widely used integrated development

environment (IDE) tailored for Python development. It is recognized for its user-
friendly interface. It provides a huge collection of tools and features, contributing to an
efficient coding process. Some of the features include:

 The visual debugging feature in PyCharm enables users to establish breakpoints,

examine variables, assess expressions, and navigate through code step-by-step
 PyCharm is equipped with integrated database tools that enable users to
establish connections with databases, execute queries, and manage databases
directly from IDE.

The Steps performed in PyCharm related to this project are:

(i) Start PyCharm by creating an appropriate file name.

(ii) Load the CSV dataset file.
(iii) Assessing the algorithm.
(iv) Visualization of results.
 8

2.2 DATASET

Dataset collected from [7]. This repository contains a dataset

(`data_set\consumes.csv`) that simulates the consumption patterns of 9 users over the
years 2018 and 2019. The dataset comprises 69,194 records, representing hourly
consumption.

It's worth noting that the original dataset is sourced from the following
repository: https://github.com/DAIAD/data/blob/master/swm_trialA.zip. The
repository includes all datasets generated during the Trials of the DAIAD project. For
detailed information about each dataset and its characteristics, please refer to the D7.3
Report Deliverable of the DAIAD project, accessible at www.daiad.eu.

All datasets come with a CC-BY-SA license.

The preprocessed dataset encompasses the following fields:

- `id` : Record Id
- `userId` : User Id
- `deviceId` : Device Id
- `time` : Consumption time (yyyy-MM-dd hh:mm:ss)
- `consume` : Consumption in liters
- `totalConsume` : Accumulated consumption
- `isAtHome` : This field indicates whether the user is at home or not,
with possible values Y (Yes), N (No), U (Unknown).
- `isAnomalous` : This field categorizes consumption as anomalous (1) or
not(0).
 9

2.3 EXPLORATORY ANALYSIS

PREPROCESSING

(I) Chosen Data:

From the given dataset containing consumption rates for 9 users, I

selected data from 5 users. I created individual datasets for each of
these 5 users, and initially, I extracted data for 10 days from the
available year-long dataset to assess my algorithm.

(II) Data Imputation:

During the preprocessing phase, I handled missing values in user

consumption by replacing them with the consumption rates recorded
at the same hour on the previous day.
 10

CHAPTER 3

PROPOSED MODEL

3.1 INTRODUCTION

Initially, our project was bifurcated into two modules. The first module focuses
on water distribution, analysing the consumption of each water tank by taking into
account various factors such as temperature, population, season, and holidays. The
second module is dedicated to managing water within nearby tanks independently until
the main water source is required.

Figure 3.1. Smart Water Distribution Network

For demand forecasting to predict the daily consumption of water, we are using
Short-Term forecasting. We plan to study the consumption rate of each water tank by
considering different factors such as temperature, population, seasons, and holidays.
 11

Figure 3.2. Water Demand Forecasting

In the initial phase, we created a water management system to address water

scarcity issues in tanks arising from excessive consumption in an area, while also
incorporating predicted water consumption.

Figure 3.3. Water Management

For the design of the water management system, we opted for the powerful and
reliable Python language. The entire management system was compartmentalized into
distinct modules, each handling specific tasks within the overall framework.
 12

3.2 MODULES

 Water Usage Management.

 Leakage Detection.
 Shortest Path.
 Controlling Valves.

3.2.1 WATER USAGE MANAGEMENT

A water management system algorithm that predicts and manages the water
consumption rates in different tanks over a specified period. Here's a report
summarizing the key components and functionalities of the code:

Summary of the Water Management System Algorithm:

(I) Libraries used:

- The code imports necessary libraries, including Pandas for handling
data, NumPy for numerical operations, and Matplotlib for data
visualization.
(II) Data Loading:
- The algorithm loads water consumption data from a CSV file named
'water_consumption.csv' using Pandas.
(III) Data Initialization:
- The data is converted to a dictionary format for ease of manipulation.
- Various dictionaries are initialized to store predicted values,
consumption rates, tank levels, water transfers, and daily/hourly
usage statistics.
(IV) Internal Water Transfer Visualization:
- The algorithm visualizes internal water transfers between tanks using
a bar chart.
 13

- Positive values represent water intake, while negative values

represent water output.
(V) Hourly Usage of Tanks Visualization:
- Hourly water usage for each tank is visualized, showing the change
in water levels over time.
(VI) Daily Water Usage Calculation:
- The algorithm calculates daily water usage for each tank, considering
initial tank levels, water transfers, and actual consumption rates.
(VII) Regulating Water Flow:
- The code contains a function to regulate water flow between tanks
based on predicted and actual consumption rates.
- It determines the amount of water needed for each tank and finds the
optimal path and valves to transfer water.
(VIII) Result and Reporting:
- The algorithm prints and visualizes the predicted values,
consumption rates, and tank levels for each hour.
- Internal water transfers and hourly tank usage are displayed for
analysis.

3.2.2 LEAKAGE DETECTION

The Leakage Detection algorithm appears to be designed for detecting leakages in

a water pipeline system. Let's break down the key components and functionalities of
the code:

(I) Libraries Used

- `pandas`: Used for data manipulation and analysis. In this script, it's
employed to read a CSV file ('piperecord1.csv') into a pandas
DataFrame.
 14

(II) Data Loading

- The CSV file 'piperecord1.csv' is read into a pandas DataFrame
named `iterator`.
(III) Data Transformation
- The DataFrame `iterator` is converted to a dictionary using the
`to_dict()` method.
(IV) Data Analysis
- The script initializes lists (`hour1` to `hour12`) to store information
about leakages for each hour.
- A list of pipes (`pipes`) is defined.
- A dictionary (`data`) is created to organize the data.
- A loop iterates through the data in `iterator` to check if the water flow
in each pipe for each hour is greater than £. If yes, it is marked as a
potential leakage ("yes"); otherwise, it is marked as no leakage ("no").
- The results are stored in the corresponding hour lists
(V) Data Output
- The collected data is organized into a new DataFrame (`df`) and saved
to a CSV file ('pipesdata.csv').
(VI) Leakage Detection
- The function `detect_leakages()` is designed to identify pipes with
potential leakages based on the conditions specified.
- The function returns a list of pipes (`leakage_pipes`) with detected
leakages and clears the `leakage_pipes` list.
(VII) Data Clearing
- After the leakages are detected and stored, the individual hour lists
are cleared for the next round of data processing.

In summary, this algorithm reads water pipeline data, analyzes it to detect

potential leakages for each hour, saves the results to a new CSV file, and identifies
pipes with detected leakages.
 15

3.2.3 SHORTEST PATH

The Shortest Path module appears to be part of a water distribution and

management system. Here's a report on the key components and functionalities:

(I) Overview
- The Algorithm focuses on managing the distribution of water among
different tanks connected by pipes. It utilizes Dijkstra's algorithm to
find the shortest path between water tanks, considering various
factors such as leakages
(II) Code Components
i. Graph Representation:
o The water distribution network is represented using a graph
with nodes as tanks and pipes connecting them.
o The `graph` dictionary holds the edge weights between tanks.
ii. Dijkstra’s Algorithm:
o The `dijsktra_algorihtm` and `dijsktra` functions implement
Dijkstra's algorithm to find the shortest path between a source
and destination tank.
o The algorithm considers leakages, adjusting edge weights for
affected pipes.
iii. Water Tanks and Pipes:
o Tanks and pipes are represented using dictionaries
(`pipes_tanks_links`,`TANKS`).
o The `find_nearest_tank` function returns a list of nearest tanks
for a given tank, considering leak adjustment.
iv. Leak Detection:
o The `leakage_detection` module is imported to identify
leakages in the water distribution network.
 16

v. Heap Operations
o Heap operations are used to efficiently manage and retrieve
minimum values during Dijkstra's algorithm execution
(III) Functionality
- The Algorithm dynamically adjusts the nearest tanks for each tank
based on leakages detected
- It calculates the shortest paths and associated costs between tanks.
- The water management system aims to optimize water distribution
considering leakages and shortest paths
(IV) Conclusion
- The Algorithm provides a foundation for a water management
system, considering factors like leakages and optimizing water
distribution. Further development and documentation could enhance
its usability and robustness.

3.2.4 CONTROLLING VALVES

This Algorithm script seems to be related to a water pipeline system, specifically

focusing on identifying the valves that need to be opened based on the pipes that are
part of a given path. Below is an analysis of the key aspects of the code:

(I) Data Structures:

- `valves_tanks_links`: A dictionary mapping pipes to pairs of valves.
- `pipes_tanks_links`: A dictionary mapping pipes to pairs of tanks.
Each entry contains two lists corresponding to the pipe direction.
(II) Functions – ‘valves_to_open(pipe_data)’:
- This function takes a list `pipe_data` representing a sequence of pipes
in a path.
- It iterates through the list, extracting pairs of consecutive pipes and
matching them to entries in `pipes_tanks_links`.
 17

- For each matched pair, it retrieves the corresponding valves from

`valves_tanks_links` and appends them to the `valves` list.
- The function returns the list of valves that need to be opened to follow
the given path.

In summary, the script provides a function to determine the valves that should
be opened based on a given path of pipes in a water pipeline system. The dictionaries
`pipes_tanks_links` and `valves_tanks_links` are crucial for mapping pipes to tanks
and valves, respectively.

3.3 COST ANALYSIS

This hypothesis outlines a scenario for cost analysis related to water distribution.
Let’s break down the key components of this hypothesis:

(I) Water Distribution System:

- Source: A reservoir is identified as the source of water.
- Destination: Water is being distributed to five regional tanks.
- Tank Capacity: Each regional tank has a capacity of 5000
gallons.
(II) Geographical Considerations:
- Distance: The average distance between the reservoir and each of
the five regional tanks is specified as 30 kilometres.
(III) Scope of Analysis:
- Cost Analysis: The primary focus is on analyzing the costs
associated with the water distribution process.
(IV) Assumptions:
- The analysis assumes certain conditions or factors that may
influence costs. These could include factors such as the
efficiency of the distribution system, the energy required for
pumping water, maintenance costs, and other relevant variables.
 18

(V) Purpose:
- The purpose of the hypothesis is likely to evaluate the economic
aspects of water distribution in this specific scenario. This could
involve understanding the operational costs, energy
consumption, and potential areas for cost optimization.
(VI) Analysis Scenario:
- This scenario sets the stage for a detailed cost analysis of the
water distribution process. It may involve considering factors
like the power requirements of pumps, energy consumption by
motors, and amount of water being saved from leakage in pipe.
(VII) Data Points:
- The hypothesis provides specific data points such as the tank
capacity, distance, and the number of tanks. These would be
crucial for performing accurate calculations and assessments.

How we are saving energy?

(I) Power required for Main Source:

Parameters:
- Total Head (H): The total vertical and horizontal distance the
water needs to be pumped.
- Flow Rate (Q): The desired volume of water to be pumped per
day.
- Efficiency (ŋ): The efficiency of the pump.
.
Power (P) = (TDH= Helevation + Hfriction + Hpressure) (1)
ŋ
 19

(II) Flow Rate in a Pipe:

Q = A.V (2)
A = cross-sectional area (𝜋 𝑟^2)
V = velocity of the water.
- The determination of velocity (V) involves applying Bernoulli's
equation or other fluid dynamics equations, and this is
contingent upon the unique characteristics of your system.
Velocity is subject to factors like pump head, pipe length, and
system resistance.
(III) To calculate the amount of energy to distribute water:
- Power(kW) = 𝐻𝑜𝑟𝑠𝑒𝑝𝑜𝑤𝑒𝑟(𝐻𝑃)𝑋0.7457 (3)
( ) ( )
- Energy (kWh) = (4)
€

Where € = Motor Efficiency ( 𝑋 100)

(IV)
 20

CHAPTER 4

RESULTS AND DISCUSSION

4.1 INTRODUCTION

In this section, our algorithm operates on a dataset comprising one year of

consumption data. During a 20-day evaluation period, the algorithm functions
seamlessly, effectively managing water internally. It adeptly addresses leakages and
autonomously adjusts pathways to divert water, thereby excluding pipes with detected
leaks.

Every 12 hours, we produce a comprehensive report for analysis purposes. This

report includes the hourly water usage in individual water tanks, the daily water usage
in a specific region, and the internal transfer of water between tanks in different regions
to address water scarcity.

For every hour the leakage database table is updated after checking for leakages
in pipe this is an hourly periodic check.
 21

4.2 PERFORMANCE

DAY 1:

Figure 4.1. Hourly water Figure 4.2. Hourly water Figure 4.3. Hourly water
consumption of Tank 1 consumption of Tank 2 consumption of Tank 3

Figure 4.4. Hourly water Figure 4.5. Hourly water

consumption of Tank 4 consumption of Tank 5

Figure 4.6. Water Usage Figure 4.7. Internal water

on date [02-11-2023] transfers on date [02-11-2023]
 22

DAY 2:

Figure 4.1. Hourly water

consumption of Tank 1

Figure 4.8. Hourly water Figure 4.9. Hourly water Figure 4.10. Hourly water
consumption of Tank 1 consumption of Tank 2 consumption of Tank 3

Figure 4.11. Hourly water Figure 4.12. Hourly water

consumption of Tank 4 consumption of Tank 5

Figure 4.13. Water usage Figure 4.14. Internal water

on date [03-12-2023] transfers on date [03-12-2023]
 23

CHAPTER 5

CONCLUSION

5.1 INTRODUCTION
In conclusion, this research aims to address the critical issue of water scarcity
through the implementation of the Internet of Things (IoT) based water management
system tailored for housing societies. Through historical data and demand forecasting,
the model guarantees that water distribution corresponds to consumption patterns. The
incorporation of detection techniques in the leakage detection module empowers the
system to identify and address leaks, thereby minimizing water wastage and
contributing to the overall sustainability of water resources. By redirecting water from
tanks with lower consumption rates instead of relying on the primary source, energy is
conserved.

5.2 HIGHLIGHTS OF WORK DONE

The integration of Internet of Things (IoT) technology forms the backbone of
the proposed water management system. Demand forecasting is done to accurately
forecast daily water demand and water is distributed accordingly. Leakage detection
algorithms are implemented to identify and address water leakages. Graph algorithms
such as Dijkstra’s algorithms is used to determine the shortest path using which the
internal transfers are performed to supply water when storage has occurred. IoT-
enabled valve controlling function is implemented to dynamically adjust the water
distribution.
 24

5.3 FUTURE WORK

 To integrate smart management module with IoT devices using cloud server.
 To implement a smart distribution module.
 To integrate both the modules.
 To develop user interface.
 To develop a small prototype using hardware components.
 25

REFERENCES

[1] Lakshmi Kanthan Narayanan, and Suresh Sankaranarayanan, “IoT-based water

demand forecasting and distribution design for smart city”, Journal of Water and
Climate Change, Vol. 11, No. 4, pp.1411–1428, December 2020.

[2] Ankit Anilkumar Maroli, Vavhav S. Narwane, Rakesh D. Raut, and Balkrishna
E. Narkhede, “Framework for the implementation of an Internet of Things
(IoT)-based water distribution and management system”, Clean Technologies
Environmental Policy, Vol. 23, pp. 271–283, January 2021.

[3] M.M. Srihari, “Intelligent water distribution and management system using the
Internet of Things”, In Proceedings of the 2018 International Conference on
Inventive Research in Computing Applications (ICIRCA), Coimbatore, India,
pp. 785-789, IEEE, July 2018.

[4] Kaushik Gupta, Mandar Kulkarni, Manas Magdum, Yash Baldawa, and
Shivprasad Patil, “Smart water management in housing societies using IoT”, In
Proceedings of the 2018 Second International Conference on Inventive
Communication and Computational Technologies (ICICCT), Coimbatore,
India, pp. 1609-1613, IEEE, April 2018.

[5] Tomas Robles, Ramon Alcarria, Diego Martín, Augusto Morales, Mariano
Navarro, Rodrigo Calero, Sofia Iglesias, and Manuel López, “An Internet of
Things-based model for smart water Management”, In Proceedings of the 2014
28th international conference on advanced information networking and
Applications workshops, Victoria, BC, Canada, pp. 821-826, IEEE, May 2014.

[6] F Jan, N Min-Allah, S Saeed, SZ Iqbal and R Ahmed, “IoT-based solutions to

monitor water level, leakage, and motor control for smart water tanks”, Water,
Vol. 14, No. 3, p. 309, January 2022.
 26

[7] H Fuentes, and D Mauricio, “Smart water consumption measurement system for
houses using IoT and cloud computing”, Environmental Monitoring and
Assessment, Vol. 192, No. 602, pp. 1-16, September 2020.

[8] M Pan, H Zhou, J Cao, Y Liu, J Hao, S Li and CH Chen, “Water level prediction
model based on GRU and CNN”, IEEE Access, Vol. 8, pp. 60090-60100, March
2020.

[9] L Wu, C Kong, X Hao and W Chen, “A short-term load forecasting method
based on GRU-CNN hybrid neural network model”, Mathematical Problems in
Engineering, Vol. 2020, Article ID 1428104, pp. 1-10, March 2021.

[10] M Sajjad, ZA Khan, A Ullah, T Hussain, W Ullah, MY Lee and SW Baik, “A

Novel CNN-GRU-Based Hybrid Approach for Short-Term Residential Load
Forecasting”, IEEE Access, Vol. 8, pp. 143759-143768, July 2020.

[11] Q Tao, F Liu, Y Li and D Sidorov, “Air Pollution Forecasting Using a Deep
Learning Model Based on 1D Convnets and Bidirectional GRU”, IEEE Access,
Vol. 7, pp. 76690-76698, June 2019.

[12] Liyuan Deng, Xiaobo Chang, and Peng Wang, “Daily Water Demand Prediction
Driven by Multi-source Data”, Procedia Computer Science, Vol. 208, pp. 128-
135, January 2022.