Implementation of Real

Time Smart AI based

Rainfall Prediction
System using Internet
of Things and Machine
Learning
 
ABSTRACT

 The system proposed in this paper is an advanced solution for monitoring the
weather conditions at a particular place and make the information visible
anywhere in the world.
 The technology behind this is Internet of Things(IoT), which is an advanced
and efficient solution for connecting the things to the internet and to connect
the entire world of things in a network.
 The data updated from the implemented system can be accessible in the
internet from anywhere in the world. Data is collected from various sensors
and uploaded to cloud servers.
 Useful information such as rainfall prediction is deduced from these data
using python. IOT technology is used to get real time parameters such as
temperature, air quality, humidity etc.
INTRODUCTION

 With the advent of high speeds Internet, more and more humans around the globe are
interconnected. Internet of Things (IoT) takes this a step further, and connects not only
humans but electronic devices which can speak amongst themselves.
 The main concept behind the Internet of Things(IoT) is to connect various electronic
devices through a network and then retrieve the data from these devices (sensors) which
can be distributed in any fashion, upload them to any cloud service where one can
analyze and process the gathered information.
 In the cloud service one can utilize these data to alert people by various means such as
using a buzzer or sending them an email or sending them an SMS etc. Existing technology
mainly focus on controlling and monitoring of different activities.
 These are increasingly emerging to reach the human needs. An efficient environmental
monitoring system is required to monitor and assess the conditions in case of exceeding
the prescribed level of parameters. Sensors are placed at different locations to collect
the data to predict the behavior of a particular area of interest.
 LITERATURE SURVEY

Year Author Methodology Conclusion

2021 K.Sarvani The project uses three algorithms We have learned different preprocessing techniques
that are required in preparing the dataset. The dataset
Y,Sai namely, Decision Tree Regression, must be free from all kinds of noise, inconsistency,
Priya ,Ch.Tej Random Forest Regression, and overfitting, and other odds that may affect the
performance of the model.
a, SVM. The project is a case study of
INDIA how the factors vary the rainfall
pattern. The prediction is
completely based on the
independent variables that cause
rainfall.
Year Author Methodology Conclusion
2021 Puja The proposed system will Devices that monitor
Sharma , work on the client-server weather’s parameters with
Shiva architecture model using minimum cost in our
Prakash, IoT. The system is proposed system. The
INDIA organized in Two-tier proposed system works on
Architecture. Our the client-side architecture
proposed system contains model. The proposed
a various sensor which approach observed various
will monitor the environmental information
temperature of the using multiple sensors. The
region, humidity, Rain system which is designed
value and pressure of the has used less sensor than
system. The sensor the existing model.
captured data and send it
to the node MCU
controller. The serial
monitors an IP address.
The HTTP protocol is
used to view the data on
the webserver.
 EXISTING SYSTEM

 There are two techniques to predict rainfall. The

first one is machine-learning approach ,which
includes LASSO regression.
 The second one is neural network approach. This
system first compares both the process and then
accordingly gives result with the best algorithm
PROPOSED SYSTEM
 In this, we present the theory on real time rainfall
monitoring and prediction using IoT.
 The system consists of ESP32 microcontroller, different type
of sensors like soil moisture sensor, air quality sensor,
Temperature & humidity sensor and rain sensor.
 The ESP32 is the main processor of the system which control
and process the data generated by the sensors. A Wi-Fi
module is inbuilt connected to the ESP32 device which help
to transfer the data to the cloud over internet.
 DHT11 measures the temperature and humidity and sends it
to the esp32 microcontroller. The data collected from all
the sensors is send to the cloud servers where it can be
stored.
 Google firebase is the cloud server used in our project.
These data is fed to the algorithm written in python idle
installed in the pc. The output of the algorithm i.e. the
prediction of rainfall, is send back to the ESP32
microcontroller to be displayed on the LCD.
 SYSTEM SPECIFICATIONS

HARDWARE USED
 ESP32 Microcontroller
 DHT11
 Rain sensor
 Soil moisture sensor
 MQ3(air quality sensor)
SOFTWARE USED
 Arduino IDE
 Embedded C
 Google Firebase Cloud
 Python IDLE
 CONCLUSION

By keeping the embedded devices in the environment for monitoring enables

self-protection to the environment. To implement this, we need to deploy the
sensor devices in the environment for collecting the data and analysis. The main
motto is to use the inexpensive components and attain the maximum best
accurate system which could monitor the weather in real time application in
agricultural lands and use this data to make more accurate future weather
predictions using python and machine learning algorithms.
  
REFERENCES

 Deshmukh A. D. & Shinde U. B. 2016, August. A low cost environment monitoring system
using raspberry Pi and arduino with Zigbee. In: Inventive Computation Technologies
(ICICT), International Conference on. 3: 1-6. IEEE.
 Savić T. & Radonjić M. 2015, November. One approach to weather station design based
on Raspberry Pi platform. In: Telecommunications Forum Telfor (TELFOR), 23rd . IEEE.
pp. 623-626.
 Wang Y. & Chi Z. 2016, July. System of Wireless Temperature and Humidity Monitoring
Based on Arduino Uno Platform. In: Instrumentation & Measurement, Computer,
Communication and Control(IMCCC), 2016 Sixth International Conference on. IEEE. pp.
770-773.
 Saini H., Thakur A., Ahuja S., Sabharwal N. & Kumar N. 2016, February. Arduino based
automatic wireless weather station with remote graphical application and alerts. In:
Signal Processing and Integrated Networks (SPIN), 2016 3rd International Conference on.
IEEE. pp. 605-609.
 Kumar N. P. &Jatoth R. K. 2015 May. Development of cloud based light
intensity monitoring system using raspberry Pi. In: Industrial Instrumentation
and Control (ICIC), 2015 International Conference on IEEE. pp. 1356- 1361.
 Srinivasan V. S., Kumar T. & Yasarapu D. K. 2016, May. Raspberry Pi and
iBeacons as environmental data monitors and the potential applications in a
growing Big Data ecosystem. In: Recent Trends in Electronics, Information &
Communication Technology (RTEICT), IEEE International Conference on. IEEE.
pp. 961-965.
 Ibrahim M., Elgamri A., Babiker S. & Mohamed A. 2015, October. Internet of
things based smart environmental monitoring using the raspberry-pi
computer. In: Digital Information Processing and Communications (ICDIPC),
2015 Fifth International Conference on. IEEE. pp. 159-16
 DESIGN PHASE

 The implemented system consists of a microcontroller (ESP32) as a main

processing unit for the entire system and all the sensor and devices can be
connected with the microcontroller.
 The sensors can be operated by the microcontroller to retrieve the data from
them and it processes the analysis with the sensor data and updates it to the
internet through Wi-Fi module connected with it.
 A Wi-Fi module is inbuilt connected to ESP32 device which help to transfer
the data to the cloud over internet.
 ESP32 can interface with other systems to provide Wi-Fi and Bluetooth
functionality through its SPI/SDIO or 12C/UART interfaces.
ESP32

 Here we have used ESP32 Wi-Fi module which is having TCP/IP protocol stack
integrated on chip.
 It uses LX6 Microprocessor
 So, that it can provide any microcontroller to get connected with Wi-Fi network.
 ESP32 is a preprogrammed SOC and any microcontroller has to communicate
with it through UART interface.
 It works with a supply voltage of 3.3v.
 The module is configured with AT commands and the microcontroller should be
programmed to send the AT commands in a required sequence to configure the
module in client mode.
 The module can be used in both client and server mode.
SENSORS
 The system consists of temperature and humidity sensor(DHT 11) and CO
sensor(MQ2).
 These 2 sensors will measure the primary environmental factors temperature,
humidity and the CO levels. All this sensors will gives the analog voltage
representing one particular weather factor.
 The microcontroller will converts this analog voltage into digital data.

TEMPERATURE SENSOR AND HUMIDITY SENSOR

Temperature Sensor & Humidity Sensor

The DHT 11 is an essential, ultra minimal effort computerized temperature and

humidity sensor.
 The main genuine drawback of this sensor is we can just get new information
from it once every 2 seconds, so when utilizing our library, sensor readings
can be up to 2 seconds old.
 It works on 3 to 5V power supply.
 Good for 20-80% humidity readings with 5% accuracy and for 0-50 degree
celcius temperature readings +or- 2 degree accuracy.

GAS SENSOR
 MQ2 gas sensor is an electronic sensor used for sensing the concentration of
gases in the air such as LPG, propane, methane, hydrogen, alcohol, smoke
and carbon monoxide.
 MQ2 gas sensor is also known as chemiresistor.
 It contains a sensing material whose resistance changes when it comes in
contact with the gas.
 It is Metal Oxide Semiconductor.
 It can detect flammable gas in a range of 300 – 1000ppm.
 The MQ2 Gas Sensor has 4 terminals.
 The terminals include the Vcc for accepting positive voltage to power the
sensor, GND for ground connections to complete the circuit and activate the
sensor.
 Dout which provides a digital output terminal and a Aout which provides the
analog output terminal.
SOIL MOISTURE SENSOR

 A resistive soil moisture sensor works by using the relationship between

electrical resistance and water content to gauge the moisture levels of the
soil.
 It connects the wires from the other (4-pin) side of the amplifier to an
Aurdino board.
 The VCC goes to 5V, GND to ground, A0 to an analog pin, and D0 to a digital
Aurdino pin.
 The run code found here to get it to output the sensor value via the serial
port.
RAIN SENSOR

 A rain sensor is a switching device activated by rainfall.

 Expansion disk rain sensors operate via a gauge attached to the sprinkler
system.
 Disk inside the gauge absorb water and expand as more rain falls.
 That sends a message to the sprinkler system controller and interrupts the
signal that turns on the sprinkler.
 The sensor acts as a variable resistance that will change status: the resistance
increases when the sensor is wet and the resistance is lower when the sensor
is dry.
 The comparator has 2 outputs connected to the rain sensor, a digital output
(0/1) and a analog output (0 to 1023).
Arduino LCD DISPLAY

 It accepts the serial input from the computer and upload the sketch to the
Arduino.
 The character will be displayed on the LCD.
THANK YOU