ABSTRACT:

The Internet of Things and Machine Learning algorithms have brought about a revolution in the field
of technology as it has treated several common problems at ease. Things have become automated
and easier due to the advent of these techniques. Floods have drained out civilizations and are a
serious threat as ever. Internet of Things along with Machine learning is tried out in this paper to
bridge the gap between arrival of the flood and the cautionary measures to prevent loss of life and
property. In this study, a unique, advanced and a very efficient IoT based sensor system is proposed
to set the scheme of sending the notification through Notification alert whenever water level and
rainfall count crosses the threshold value. The other extent of this paper is the future prediction of
upcoming floods and heavy rainfall through Machine Learning algorithm.

INTRODUCTION

Situation of flood victims and henceforth, it reduces the disaster effect. The alarm monitoring is a
vital task in complex industrial services like power systems [1], chemical industries [2], oil refineries
[3] and large-scale networks like computer [4], sensor networks [2] and communication [6]. These
are generally controlled with a large number of communication devices, sensors, control modules
and actuators. It is a great challenge for designing an efficient alarm system considering the complex
features of these components, such as nonlinearities, process correlations etc. Likewise, in this
paper we deploy IoT to send information from the sensors connected to Node MCU to the
Thingspeak server. This data is used to generate SMS notification whenever there is a fluctuation
beyond threshold level. Furthermore, we use Machine Learning algorithm ARIMA to predict the
occurrence of the upcoming floods. Flooding has become one of the major issues in India. It has
become a disaster in many places of India leading to loss of life, time, property and money. As such
India is one of the world’s most flood prone countries today. One of the major causes for heavy
rainfall is the South West Monsoon winds hitting India in the monsoon months. Human beings suffer
much due to the showers since they are not aware of the upcoming events. An alerting and future
predicting system cannot prevent floods but can help to save people, property or money. Internet of
things furthermore referred to as the commercial net has been able to successfully connect billions
of devices together. It is expected that 50 billion devices are supposed to be connected by the end of
2020 through IoT technology. IoT coverages the advancement of various arenas like industry,
agriculture, healthcare and automobiles. It also provides improved home automation along with
automating numerous routine activities

EXISTING SYSTEMS
There are various existing flood detection and smart alert systems currently in use, with different
levels of sophistication and effectiveness. Some of the existing systems are:

1. Traditional flood monitoring systems: These systems use the manual measurement of water levels
and flow rates through traditional techniques like staff gauges, weirs, and flumes. However, these
systems can be timeconsuming, labor-intensive, and prone to errors.

2. Remote sensing and satellite-based systems: These systems use remote sensing and satellite-
based technology to monitor water levels and flow rates, and detect changes in land use and
vegetation cover that may impact flood risk. However, these systems can be expensive and may
have limited spatial and temporal resolution.

3. IoT-based systems: These systems use a network of sensors and devices to collect and transmit
real-time data on water levels, flow rates, rainfall, and other relevant parameters. These systems can
provide real-time information but may require a significant investment in hardware and
infrastructure.

Proposed System:

 Nowadays, there is no idea about when flood will occur, so there is need to prepare such
system for people who are near the flood prone area. Therefore, the purpose of this
system's design is to send out alert signals to individuals about the approaching flood.. The
basic working of project consists monitoring of water levels and values of other sensors used
against predefined threshold values and if collected sensor values exceed threshold values
potential flood is detected and alerts are sent through Notification. After thorough literature
survey we decided to use
 DHT11 (Temperature and Humidity Sensor) to detect real-time temperature and humidity of
environment.
 HC-SR04 (Ultrasonic Sensor) to check rise or fall in water level.Once the water level increases
beyond threshold, a trigger is generated that causes an SMS Alert warning about the
potential for flood.
 Ultrasonic Sensor to check real-time water flow.
 Buzzer to give alert when it gets triggered.
 For software setup we have Nodemcu and Arduino IDLE for writing Embedded C languages
program. The Nodemcu is used by the application to read the sensor inputs. We set a
 threshold value in the ultrasonic sensor program for the water level in three levels: normal
water level causes the green LED to turn on; at medium water level, the yellow LED glows
on; and once the water level reaches the highest threshold level, the red LED and buzzer
turn on. In other words, if the water level is at the closest distance from the ultrasonic
sensor, a trigger instructs the program to send out an Notification warning people about the
potential for flooding and rising water levels. We have also written program for rain sensor
and water flow sensor for getting Notification regarding Rain and high water flow
respectively.Furthermore, we also created channel on Arduino Cloud Platform for displaying
real-time sensor values in graphical form which makes it easier to study sensor values time
to time.

BLOCK DIAGRAM

Power
DH11 Sensor
Supply

Ultrasonic
NODEMCU
Sensor Red Led
MICROCONTROLLER

Buzzer
Green Led

Cloud
Mobile
Phone
or

Laptop
HARDWARE COMPONENTS

 NodeMCU Microcontroller
 DHT 11 Sensor
 Ultrasonic Sensor
 Green led
 Red led
 Power Supply
 Buzzer

SOFTWARE COMPONENTS

 Arduino ide Tools