AgriCol

AgriCol is a AI based helping-hand of farmers - helps them to increase yield by suggesting right crop at right time.

Comment

Inspiration

The global population is expected to reach 10 billion people by 2050,which means double agricultural production in order to meet food demands which is about 70% increase in food production.Farm enterprises require new innovative technologies to face and overcome these challenges.AI helps to resolve these problems. The knowledge and infrastructural gaps, particularly in rural regions, are the major problems affecting Indian agriculture today. Infrastructure issues with regard to markets, transportation, and irrigation all significantly increase the cost of farming. The absence of delivery systems is another problem. There are several programmes designed to advance agriculture. In terms of boosting productivity, cutting costs, or improving price realisation on the ground, we lack efficient delivery systems that may convert into efficient facilitation. Crop planning is also an other important problem in agriculture sector. That is, to select proper set of crops to be cultivated at particular season for maximization of yield and minimization of crop wastage.A crop plan is complete only when it contains the following factors,such as pH of soil,temperature,humidity,climatic details,etc..The existing crop planning application does not met the desired accuracy level.AgriCol comes here with a solution.

What it does

Agricultural systems pose many challenges and problems that can be formulated as optimization problems.Among these one of the major challenge is crop selection. That is, to decide the proper set of crops to be cultivated with proper irrigation scheme. Such decisions are made for the maximization of net profit and minimization of crop wastage. Proposed Methodology: AgriCol is an AI based guiding application that provides a complete crop plan to farmers(sowing to harvesting) using soil data(soil type,pH) and environmental factors such as temperature, humidity,etc.. The crop plan consists of: Success rate of crops at particular season Days of germination and maturation Nutrients to be supplied Irrigation and pest prevention information Accuracy of AgriCol is more than 90%.Farmers can get all details about the crop in a single application and finally increases yield.

How we built it

AgriCol is an MERN based application along with deep learning connectivity using flask.

Data Used:

For this project, we have collected some of the data from farmers such as pH,N,P,K values,soil type, water availability, irrigation details,location details. We also use rainfall,temperature,humidity data from farmers.

Backend Model:

The backend model consists of three different models combined together to increase the accuracy.Initially using the NPK and pH values the model provides an output with approximately 10 values.This along with soil type and water availability in the second model alters the crop with maximum probability of success. To improve the accuracy further we have also passed the output with crop yield data to get high accuracy.

Challenges we ran into

As it consists of a large tech stack, we found it difficult to get project done in short span of time. It took time for us to learn the stuffs.Further we have challenges in connecting python flask and reactjs. It took some time to learn and connect it.We found some difficulty in connecting third party API along with our code.

Overall it was a nice learning experience through Turtle Hacks. Thanks for the opportunity.

Accomplishments that we're proud of

We have made a co-ordination between our team to work on this project and

What's Next for AgriCol

The development phase is almost done. We have planned to take this furthermore. So testing is to be done with some of the manual and practical data to test the accuracy. Next we must get clarity with farmers for our project. Mostly we also need regional language so that farmers can use our application wisely. In the deployment phase we will be using AWS to host our product to the outer world!!!!

What's next for Untitled

Share this project:

Updates