IOT Based Smart Farming for Precision Agriculture

Agriculture and farming is considered the backbone of Pakistan's economy, which relies heavily on its major crops. Agriculture accounts for about 18.9% of Pakistan's GDP and employs about 42.3% of the labour force.(Ministry of finanace)   Now when the environmental crisis are beco

Project Title

IOT Based Smart Farming for Precision Agriculture

Project Area of Specialization

Electrical/Electronic Engineering

Project Summary

Agriculture and farming is considered the backbone of Pakistan's economy, which relies heavily on its major crops. Agriculture accounts for about 18.9% of Pakistan's GDP and employs about 42.3% of the labour force.(Ministry of finanace)

Now when the environmental crisis are becoming large day by day there is a need of smart farms and precision agriculture in order to enhance our crop production smartly using latest technologies for betterment of mankind.

Smart Farming includes the totally automated and manged farming techniques using different types of sensors those were linked together through Wireless Sensor Networks.Further it is fully capable of detecting of pests in the crops and alarms the farmer about them so he can take necessary measures according to it.

Smart Farming include smart irrigation that efficiently overcomes the problem of low water resources and also it includes intelligent usage of environmental conditions helped the system to intelligently irrigate the farm according to need efficiently.

Precision agriculture seeks to use new technologies to increase crop yields while lowering the levels of traditional inputs needed to grow crops (land, water, fertilizer, herbicides and insecticides). It uses smart Soil moisture sensors to actually detect the physical condition of earth that whether it has a need of water or not that is included in a complete mesh which is the actual physical system installed in a field to detect physical condition and send it to main controller for intelligent smart irrigation, In our project it comprises of two meshes those are linked using Wireless Sensor Networks Technology at these Cyber Physical Systems. Another system working in parallel to it is smart weather conditions fetching system that fetch the latest and accurate( using latitudes and longitudes) weather data from Internet(Google) at that place.This data includes the rainfall,wind speed, perception of rain in next four days and humidity rate in the atmosphere.At main controller this weather data fetched from Google and actual data received from Meshes through Zigbee Wireless Technology is further compared and  smart final decision of starting water pump is made.

Precise agriculture system also includes a complete system of pest detection which involves detection of pest and after successful detection it automatically alarms so the pest around the field repelled from entering into the field. The process is done by using Machine Learning process. It uses camera as hardware input tool from which, real-time video frames are acquired. Then these frames are sent to main controller where the detection process is done by using already trained models.

Project Objectives

1. Dynamic and Unmanned Irrigation system which reduces the wastage of water.
2. Reduction in Man power and expenditures to grow the crops using this fully automated unmanned System.
3. Linking the whole system with weather forecasting to efficiently manage the use of water in the ecosystem by comparing this data with actual physical data received from fields..
4. Automatic and precise detection of Pests in the crops through Machine learning to efficiently detect pests in the crop and make an alarm about the detection.
5. Usage of Wireless Sensor Network (WSN) and Zigbee Communication Technology for better communication between different subsystems such as smart irrigation and pest detection.

Project Implementation Method

Project implementation method includes following points.

Design of Mesh System such that there are total two meshes present and each mesh consist of Soil moisture sensor that receives data from soil and after it is linked with microcontroller to convert this analogue data in to the digital and send it to Zigbee module from where it will be communicated to another mesh or main controller.Similarly another Mesh with same components is also present in another part of field from where it collect physical earth conditions and through Zigbee wireless technology communicates the data to other Mesh or Main Controller.

As these Meshes are mutually connected through Zigbee Wireless Communication Technology so each Mesh collects physical earth data(Moisture,humidity) from different parts of farm and send them to Maincontroller.So that Earth Physical Conditions(Moisture, Humidity) can be received from different parts of Earth.

MainController comprises of a Raspberry Pi connected with internet collects the weather forecasting data from the google using the longitude and latitude of particular place and using built in API system that fetch weather conditions from a weather forecasting website on regular interval basis after every four hours.At the same time Physical Earth data(moisture,humidity) from mesh is received at Main Controller and then these two Conditions one from physical earth and other from Online Weather Forecasting is compared and decides that earth has a need of water as per its physical condition but keeping in view the weather forecasting data it process that is there  chance of rain or not? What is speed of wind at that place?, it decides to turn on Sprinkler when physical conditions demand irrigation of soil and there is no forecast of rain, if there’s a condition that earth demands irrigation but there is forecast of rain in next four hours it does not turn on Sprinkler.

For pest detection, a camera is installed on the crop field, which gets real-time video of crop field. The camera is connected to main controller. At main controller, we have already trained model for pest detection which analyze the video and run the detection process. After a successful detection, the main controller automatically turn on the alarm for few minutes. In this way, this system constantly detects the pest.

Benefits of the Project

It provides following benefits for the mankind and Society

1. Smart Irrigation System by using Wireless Sensor Network can provide a fully automated earth physical conditions like moisture level or humidity more precisely and accurately.
2. Zigbee Technology is used for wireless communication between Meshes and MainController that has an edge of less power consumption over other wireless communication technologies like GSM,WIFI and Bluetooth. Further this technology is most feasible and good to communicate this type of data over distances as Bluetooth can not transmit data over wide distance while WIFI and GSM technology use more power and there is also a lot of un-allocated bandwidth remains in them where no data communication takes place.
3. Weather Data fetching from the Weather forecasting website using API methods on regular intervals make the farmer up to date about the weather conditions on its field and as this weather forecasting system is integrated with the Mesh system that collects actual physical earth conditions and then in Main Controller they are further compared in order to turn on Sprinkler or not.So this system will make a beneficial use of weather data in irrigation and prohibits the irrigation when there is a forecast of rain, similarly it also examine other weather conditions to make a better decision for irrigation.
4. The Pest Detection will timely detects the presence of pests those are very dangerous to crops production and alarm the farmer to remove them, in such a way a very precise and much accurate way of pest detection can be done that can detect the pest in minutes using machine learning based trained model already installed at main controller with much accuracy.So this smart Pest Detection will decrease the chance of crops loss due to pest and thus Crops Production will increase.
5. All this system is fully automated without an aid of person that reduces the man resource for growing large number of crops, also this smart system make it easy for farmer to grow large number of crops now without the headache of its Irrigation or Pest detection and that will increase the use of uncultivated land to grow crops on them and such crop production and be increased boostly. Further the use of this precise and smart agriculture system enhance the crop production by smart irrigation at proper time and pest detection accurately and alarming in time as it is generally be seen that irrigation of crop is more often get late or they are irrigated when there is no need of it so this system will make it smart and whole irrigation system become more precise and accurate.Similarly Pest detection in crops is observed after some time when pests are increased in numbers in crops and it is really tough for farmer to examine the existence of pests in crops on regular basis,So this system will make this problem solved as pests are automatically detected as soon as possible as they fly in field without the need of farmer in much fast time.

Technical Details of Final Deliverable

Final Deliverable of Smart Irrigation for Precise Agriculture consist of three systems integrated together whose details are as under:

Physical Earth Conditions Data Fetching:

This system is also called MESH that comprises of a soil moisture sensor that collects data(moisture level) from the earth and sends it to MicroController ATMEGA 328 UNO BOARD where this analogue data is converted to digital form.Now the ZIGBEE MODULE is also connected with this MicroController through ZIGBEE SHEILD that communicate this data between different MESHES and also to MAINCONTROLLER where further Data Manipulation Takes place.

Main Controller:

Main Controller mainly comprises of a RASPBERRY PI connected with internet and a LINUX Operating System Installed on it. Through Internet in operating system it collects weather forecasting data from the website using its API.Its demands the Longitude and Latitude of the particular area for which we have to collect weather data and Rain Threshold for which it sends the signal of rain for data comparison. It is also connected to ZIGBEE MODULE through ZIGBEE SHEILD that collects data to be transmited from MESH and stores this data for comparison. Now we have the data from both of our points and its is manipulated further that rather its a need of irrigation or not keeping in mind the physical condition of earth and weather forecast at that time.If earth condition demands irrigation and there is forecast of no rain or less rain then our threshold at that time it turns on the sprinkler for irrigation.

At main controller there is a program that keeps on running all the time which check the weather forecast automatically on regular base after four hours and keep on sending this data for manipulation while the Physical Data collection is keep on going all the time.

PEST DETCTION:

For pest detection, we have used Machine Learning. A camera (USB/wireless) of 48 mega pixel is installed on the crop field, which gets real-time video of crop field. The pest under 5 meters range can be detected. The camera is connected to main controller. Trained model is obtained by first creating datasets and then through python programming, yoloV5 algorithm is applied to make a trained model. At main controller, we have already trained model for pest detection which analyze the video and run the detection process. After a successful detection, the main controller automatically turn on the alarm for few minutes. In this way, this system constantly detects the pest.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Agriculture

Other Industries

Others

Core Technology

Artificial Intelligence(AI)

Other Technologies

Internet of Things (IoT), Others

Sustainable Development Goals

Zero Hunger, Sustainable Cities and Communities, Responsible Consumption and Production

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com