Multilingual Agriculture Advisory System

Agriculture Advisory System

CLIENT OVERVIEW

Our client is a 50-year-old non-profit international agricultural research and training organization focusing on cereal grains and related cropping systems. The organization’s research has helped tens of millions of farmers grow more nutritious, resilient and productive crops, using methods that nourish the environment and combat climate change. They needed an agriculture advisory system.

They collaborate with hundreds of partners in 50 countries and are part of the global network of CGIAR research centers, delivering science for a food-secure future.

PROBLEM

Our client is a non-profit government organization that partners with the Department of Agricultural Extension (DAE) and Sub Assistant Agricultural Officers (SAAO) in Bangladesh. The client provides advice to farmers on the appropriate crops to grow according to Bangladeshi weather conditions such that crop wastage owing to natural disasters is mitigated.

The SAAOs work as field agents in collaboration with the DAE, to educate the farmers on a daily basis based on the weather and crop information pertinent to their region or upazila.

Our client was using an existing agriculture advisor application, which was slow, inaccurate, and heavyweight. The client wanted to rebuild the entire application and accommodate more accurate weather and crop advisory, language localization, and easy-to-use interface. Thus, an integrated agriculture advisory system.

CHALLENGES

The application required the SAAOs to login to use it. The onboarding of 14,000 SAAOs into the application posed a major challenge. In addition, establishing a communication channel to engage with the SAAOs and intimate them about the launch of the application posed another challenge.

The weather forecast data is obtained from the Bangladesh Meteorological Department (BMD) every three hours. This data has to be processed every 15 mins, including critical information such as minimum and maximum temperature, and so on. This requires a lot of time and processing effort consumption, making the application a heavyweight application and slowing down its performance. High Peak had to figure out a way to make the application lightweight and ensure that the system performance wasn’t affected.

The existing application was coded in R. The development team had to migrate the application to Java to facilitate faster and less complicated processing.

The weather data obtained from the BMD covered 140*140 grid points which wasn’t enough to obtain accurate weather conditions for smaller locations, also called upazilas. To obtain more accuracy on weather data for smaller locations, the team used an algorithm that interpolated location data.

SOLUTIONS

High Peak designed and developed a responsive web agriculture advisory system for agricultural agents in Bangladesh. This application, available both in English and Bangla, aids SAAOs in accessing relevant weather information for a five-day forecast for select crops in every upazila.

The intelligent agriculture advisory application includes:

Weather prediction system

This intelligent agricultural advisory application includes a complex weather prediction logic that takes into consideration the active weather conditions, temperature, rainfall occurrences, humidity, moisture levels, and so on, and predicts the weather in the near future.

This prediction logic is highly accurate, considering that it was built on the data obtained from the BMD and then interpolated for locations smaller than the 140*140 grids that the BMD provides. This means that, the smaller the locale focus, the more accurate the weather prediction.

According to these predicted conditions, advisories are built up with the help of agricultural experts and are stored in the advisory management system.

Simulation of real-time geographical view

This application offers a geographical view of Bangladesh that can be zoomed into focus on a particular region, district or upazila of choice.

The user can easily access the crop and weather data for that specific location by simply clicking on the chosen area. The geographical view also displays alerts in the form of pointers where advisories have been updated.

The user can also bookmark locations to save corresponding data by using the ‘Starred Locations’ feature for faster and easier information retrieval at a later stage.

Crop management system

The application includes a smart crop management system that catalogs many kinds of crops that are indigenous to Bangladesh and those that can potentially be nurtured under Bangladeshi weather and soil conditions.

This information also includes important properties of these crops, such as their ideal climatic conditions, water intake level, and so on.

Using this data and accurate prediction of weather conditions, the application provides useful crop advisories to mitigate crop damage.

Advisory management system

The advisory management system is another important part of the application. Based on present and predicted weather conditions, advisories are provided for each type of crop available in the crop management system.

These advisories are available on the platform as well as sent to the SAAOs in the form of detailed automated emails.

The SAAOs use these advisories to relay important information to Bangladeshi farmers and advise them on the most suitable crops to grow in the given climatic conditions.

User management system

The application is built for users spanning upwards of 14,000 in number. The two main users here are the DAE and SAAO agents.

The guided tour feature facilitates onboarding of the SAAO agents into the application allowing them to understand and learn navigation through the application.

TECHNOLOGIES USED

Frontend

  • React.js

Backend

  • Java

Database

  • MongoDB

Hosted on

  • Google Cloud Platform

Subscribe to our monthly newsletter