Technology in Agriculture

An overview of AI and ML applications in the field of agriculture

The world’s population is expected to grow to almost 10 billion by 2050, boosting agricultural demand. This is however challenged by the loss of biodiversity, spread of pests and diseases of plants and animals, and most importantly, the challenge of agriculture competing with other industries for shared natural resources such as water and land.

Hence the question arises can we achieve the required production increases, even as the pressures on already scarce land and water resources and the negative impacts of climate change intensify?

We believe we can and technology is the enabler of this. We are living in a time where there is a technology revolution going in the agriculture sector.

In the last months edition — Sensors for Agricultural Utility, we deep dived into the sensors that are available in the market to gauge various atmospheric factors that impact the plant growth. But sensors alone does not provide much information of decision support to the farmers. In this edition, we jump into the Analytics, Artificial Intelligence (AI) and Machine Learning (ML) space, which helps bring meaning to the sensor data. We also explore other technologies that have penetrated this market.

APPLICATIONS:

AI and ML is and can be used in a number of applications in the field of agriculture. Some are mentioned below:

1. Water and Crop management
2. Yield Prediction
3. Crop Quality
4. Disease detection

Water and Crop Management

There are multiple factors, such as soil, weather, rainfall that impact the level of water that is required for a crop. The most commonly used method is the estimation of the hydrological cycle, using studies such as the water balance, that can help design effective irrigation systems. In simple terms water balance is nothing but a math that the amount of water lost by the plant and the soil surface, is the amount of water replenishment that the plat requires.

Trending AI Articles:

1. Natural Language Generation:
The Commercial State of the Art in 2020

2. This Entire Article Was Written by Open AI’s GPT2

3. Learning To Classify Images Without Labels

4. Becoming a Data Scientist, Data Analyst, Financial Analyst and Research Analyst

This simple analysis is done with the help of knowing various factors such as soil water holding capacity, the rainfall in the farm, and irrigation that has been provided by the farmer previously. In addition factors such as temperature and humidity can help understanding the water loss in the atmosphere.

Once the micro climatic conditions are know for the farm, it is quite easy to know the hydrological cycle. This can help in predicting daily water requirements for the farmers, and additionally help them reduce water usage, while improving the yield.

Additionally, while there are multiple ML models such as linear regression, Random Forrest, and Artificial Neural Networks (ANN) can be used to predict the future water requirements. But due to the ANNs ability to simulate non-linearity among the interacting factors in the systems, it is generally considered to be a great model for predicting water requirements for crops, based on weather parameters.

Fig 1: Neural Network Representation

2. Yield Prediction

Prediction of crop yield mainly strategic plants such as wheat, corn, rice has always been an interesting research area to agro meteorologists, as it is important in national and international economic programming. There are a lot of yield prediction models that have made and are generally classified into two groups a) Statistical Models, b) Crop Simulation Models. But recently, application of Artificial Intelligence (AI), such as Artificial Neural Networks (ANNs), Fuzzy Systems and Genetic Algorithm has shown more efficiency in solving the problem.

Big Data Jobs

3. Crop Quality

One of the major challenges in Indian Agri eco-system has been a quick check to understand the crop quality, sizes, shapes and defects, without taking the effort to travel to far off farms. On the other hand farmers are also challenged, and end up travelling long distances to sell the produce, only to be rejected by buyers, since either the sizes requirements or quality don’t match the demands.

A lot of work is being done in the field of image processing (Computer Vision) to get some parameters of crop quality. Specifically, CV is used for classifying food products into specific grades, detecting defects, and estimating properties such as colour, shape, size and defects.

Fig 2: Representative image of using CV to detect defects in Tomato

4. Disease Detection

Many food producers are struggling to manage threats to their crops like diseases and pests, made worse by climate change, micro cropping and widespread pesticide use.

To overcome this many AI based systems are being developed, which can look at photographs and tell what disease the plant has, and also give recommendations and treatment options to farmers. The technology usage is quite similar to the one used in assessing crop quality. Image processing is being done to correctly identify various diseases and give recommendations to farmers. While this type of detection is reactive in nature (i.e. done only after the pest/ diseases infestations has already occurred). There are studies being conducted on providing preventive warning systems to farmers based on weather conditions. This kind of preventive warning for pests are disease are well documented for grapes and widely used.

Sources of raw data

Lastly, for using any models, it is important to have good and reliable data sources. Some of the sources are :

1. Weather Data — IMD provides historic weather data from multiple stations across India. There are also other free resources that provide free weather and rainfall information. For microclimatic conditions, having a weather station on the farm can help with better analytics

2. Other types of data — data.gov.in, data.world and Kaggle are great resources to get started.

We have only scratched the surface in using these technologies towards agriculture. There is a huge scope for these technologies to help us grow sustainably and meet the future demands.

We at GramworkX help in precision farming including integrating field data, weather patterns to drive agronomic advice to farmers and yield forecasting. We are building smart products at affordable prices for the farmers for a sustainable tomorrow. This company was born from the desire to be ready for an agricultural transformation which has its core values at poverty reduction, food security and improved nutrition. Our solution helps in quantifying and providing analytical insights into water consumption patterns across fields and soil types and providing data support systems into the amount of water required for irrigation. We aim to bring predictability to farming.

Don’t forget to give us your 👏 !