Water Management And Distribution Control Using Soil Moisture Sensor
Abstract
Development of sensor technology makes human life easier. We have soil moisture sensor to sense water content in soil and tell us the nature of soil. By the use of this sensor we can make our farming smart. In current world water conservation is major problem whether it is for drinking or irrigation. We will be controlling water by distribution by this sensor for water conservation which will be usable at time of un-availability of water for irrigation or daily use. By this sensor we will give required amount of water and avoid wastage. Our sensor is moveable but not wireless but in future we can convert it to a wireless sensor. Results we got from this sensor was not linear this was due to its behavior of sensor in soil which was very sensitive. We took ADC values for different soil samples and after regression analysis we got a linear line with all the data points around it. Our results were satisfying at the end as we were able to get zero percent moisture from sample soil. Our results are usable as a reference values.
Motivation
Major occupation in Pakistan is agriculture. In Baluchistan there is lack of agricultural land but still water is needed for daily use and main source of consumption of water is well digging. Though some parts have karez system but if we talk about main city Quetta people use to bore for water in houses and some peoples selling water having their own bores. Digging water out is causing Quetta drown by 1 cm every year.
As Baluchistan is second largest fault line of this world so, this digging of water may cause incident like earthquake of 1935 and 2008. If we can control our water distribution in gardens, crops and plants then we will be able to avoid water wastage which means the digging of water will be less as compared to wastage of water where digging was higher. By this project we will be able of water distribution to plants and gardens especially in irrigation zones where, if we control water distribution then we will have better crop in term of yield. In Baluchistan controlling water distribution is for irrigation zones where farming is manual and by this project we will measure water and then on need, we could water plants gardens and crop fields. But the question arise here is “When to water plants, crops, or gardens? ” for this purpose we need to know about the nature object plant, crops and gardens this is well explained in. For better crops and nourishment of plants we have to give complete water based on requirement depend upon nature of plants, crops and soil.
Introduction
Soil Moisture Sensor can improve decisions of irrigation such that how much water to be given and when? [5]. In this project we are using soil moisture sensor (Hygrometer)[6], which work on capacitive and resistive phenomenon. Soil moisture sensor can give us both digital and analog value. Two probes have conductive material plated on it when we insert it into soil its resistance changes and sent signal to Arduino. Arduino give us analog value and then later on we convert that value to percentage to be displayed on Nokia 5110 LCD. Arduino Uno act as central unit of whole process Arduino turn Sprinkler on if the value is below the threshold value according to the readings from soil moisture sensor. We set that threshold values after performing experiment for sample soil. Farming system will be somehow smart after this project as we will move on to smart system. This sensor will be helpful in water conservation which is major requirement now a days. By starting use of this sensor in daily life water usage for gardens we will conserve water easily.
Block Diagram
Initially when we start project our sensor starts taking readings and sent signals to Arduino so that Arduino could decide to start sprinkler or no. LCD is used to print the percent value of reading taken from sensor the percent value will be our recorded moisture of soil.
State of the Art
Current state of the art for this project is the group worked on it is using Arduino Uno and Raspberry Pi in this regard and their project is plants etc. When moisture level gets down of threshold and temperature was at low value then Arduino will on water to soil. They got their threshold values after examination of plants and their soil. In their project Raspberry Pi monitor and keeps data of soil moisture. This system is related to automation which can be controlled someone from any location this is convenient system as compared to manual farming. Its heterogeneity is high due the use of Raspberry Pi only to store data.
Problem Statement
In the project above there are few problems like they are using more than one board to keep record of soil moisture which is not needed. The project is for plants and gardens so on higher level such as crops fields they don’t have conditions to use this sensor. They project is working all on analog values so they don’t have any medium for digital values such that screen, LCD. If sensor is out of order some time than it can lead to over water in some sensitive plants and garden grass. As the group working on project monitor all process from one place so their sensor is still and not moveable so they need more than one sensor in small area. The project was having effect of temperature and weather condition on sensor. We were facing sensitivity problem continuously in taking reading soil moisture sensor is very sensitive to we and moisture soil even if we press it in hands it gives us result. To overcome this factor we put our sensor in a still place and we move our beaker to it rather than moving sensor into beakers.
Objectives
Smart farming is better than manual farming system. Considering our country where is there is deficiency of water the solution of this problem is we have to control water distribution so we are calibrating soil moisture sensor. We are using only Arduino which is central unit of whole project and soil moisture sensor which will sense soil for its level of moisture. In addition to these we are using Nokia 5110 LCD which will show us exact percentage of moisture in water. This project is low in cost and efficient to sense moisture for any type of soil, grass, crops and plants. We will not waste water now while watering grass and crops. Our sensor is directly connected Arduino so we can move it easily to any part of crop field we only have to install sprinkler in field such that each sprinkler could water given area. Our sensor has no effect of any weather condition and temperature. Initially we got some environmental disturbance but later on we get rid of it by change coding and place of reading.
Methodology
Initially we start our experiment after mixing up soil and water for valid experiment and then we penetrate our sensor probes into beaker where we kept our mixed soil for experiment. After placing probes we wait for few seconds so that sensor could adjust its reading. These probes sent signals to Arduino Uno and Arduino convert these signals into analog value which according coding we convert into percentage and then converted value will be displayed on LCD. That will be our final reading for each beaker. We will take three readings from each beaker and then we will take average for each beaker so that we can avoid error in one beaker.
Soil Moisture Sesor
Different types of sensor can be used for measuring moisture in soil. We are using soil moisture sensor. This sensor has two probes and conductive strip is plated on these probes. This sensor can give both analog and digital output. Output from sensor is input to Arduino.
Amplifier circuit use to process signals of sensor toward Arduino. Amplifier has L393 comparator IC with a potentiometer. We use to read ADC values from sensor and display it on Nokia 5110 LCD. According to ADC values Sprinkler will be on or off depending of conditions we set.
Arduino Uno
Arduino Uno is a central unit of whole project. Arduino handle switching of sprinkler as well it gets analog values from sensor and using inbuilt ADC (analog to digital converter), values retrieved and stores in variables. Arduino convert analog values into percent value and display on LCD.
Data Acquisition
We have performed four different experiments in different conditions for data acquisition and characterization record, readings from each experiment. We did three experiments for each data point so that our result will be accurate and satisfactory. After performing all the experiments we plot all the reading into MATLAB to check the linearity of graph of readings we took while experimenting with wet soil. At some points graph was a little linear but we also got sudden jumps in graph this was due to sensitivity of sensor. For best fit line we performed regression analysis within MATLAB that shows us linear line while data points in graph were around that line which was linear graph of our experiment. The linearity of this sensor was not proved.
Results and Discussions
We have performed four experiments in total in which our first two experiments we were using only water in calculated amount which 2 ml in first beaker and then with difference of 2 ml in next four beakers while our soil was not calculated we took rough measurement of soil. Result from first two experiments were not linear infect it shows sudden jump at different data points. While in other two experiments we were using calculated amount of water as well as we were using calculated amount of soil. We perform third experiment using Soil and Geo Tech Laboratory in BUITEMS and fourth experiment was performed in Control Automation and Robotic Laboratory (CARL) BUITEMS in order to have different conditions. Following is table of difference of soil and water ratio in all experiments.
Moisture Detection
While performing first experiment for moisture detection we did not calculate soil but were using calculated water. In first experiment we found that there is no linearity in graph taken from MATLAB. This is due to the calculation error as we were not using calculated amount of soil. When we were performing second experiment we were again not using calculated amount soil so we got less accurate linear graph from MATLAB. This time we got little command over graph while we were taking reading we make our sensor probes little stable so that we can avoid pressure on sensor. For the third experiment we had calculated amount of water as well as soil which was about 80 grams of soil per beaker and 2ml of water in first beaker and with difference of 2ml at each beaker. At this stage we got linear graph but with a jump where our sensor give us much higher value then last one.
We can say it might happen due to mixing of water and soil as we were mixing it manually. Our last experiment was performed in Control Automation Laboratory where conditions was completely changed then that of Soil and Geo Tech Laboratory. This time our result was again not linear but result was totally different from last three experiments this change may be due to environment we changed it. And it could be due to the amount of soil we increased. In future we can eliminate all these errors. Talking of overall result we did not get a linear behavior of sensor. This is due to its sensitivity and behavior to moisture. The graph below includes all the results we got from all four experiments. The graph shows all four experiments with name reading 1, 2, 3 and 4. All the reading was taking after taking average for each data point during experiments. This enables us to have accurate reading for each data points we were dealing.
Regression Analysis
For all the experiment we at last apply regression analysis so that we can get best fit line for all our experiments linear line is best fit line in the following graph while data points are all around the line which are our reading.
Future Work
In future we can make this sensor wireless using same sensor, Arduino Uno and Bluetooth module. As we will be putting sensor into soil first to sense moisture and then sensor value will decide whether to water crop field or no. So we will install Bluetooth module into sprinkler with some electronic circuit so that Bluetooth can get data from sensor wirelessly and then circuitry will be able to decide whether to give water or no. Another addition to this project is we can include temperature and humidity sensor to make its reading more accurate reading and make it sure that sensor is completely free from any effect such that environmental effect, temperature effect (if it has effect on sensor).
Conclusion
In Pakistan farming is manual as compared to developed countries. We took a step toward smart farming where at start we will be able to water crop fields, garden, and plants. This sensor enable us to do some Engineering in real time we do got some problems in start but with passage of time we overcome these issues and now we are able do to it again and again for other daily life projects like this one. After going through this project we come to know about some facts that was really shocking for our future we may not be able to stop incident that are going to be happen in future but if do some Engineering than we might be able to extend the happening age of these expected incidents.
