Smart detection of rice disease and insects through 4IR technologies

Barishal

Bangladesh is embracing Artificial intelligence (AI) for the digitalization of the nation. AI is the simulation of human intelligence processes by machines, especially computer systems. The high-yielding varieties and agronomic good practice play a vital role through modern technologies like machine learning method (MLM), artificial intelligence (AI) and big data which can create a dynamic shift towards modernizing agriculture. The crop and soil monitoring system will be developed through AI application. Particular applications of AI include expert systems, speech recognition & machine vision. According to work plan 8.4.3 activity of National ICT Policy-2018 has stated that it will develop a single and complete agricultural input and cropping plan with integrated advisory service through IoT, Sensor, AI, Big Data analytics and real-time data feeding system. Many activities are ongoing based on technology such as digital agriculture platform generating base data for AI, IoT for crop field condition, crop stage mapping using satellite image processing and vertical agriculture for prediction mapping (National Strategy for AI Bangladesh 2019-2024). So, BRRI has taken initiative to develop a sensor-based rice pest management ‘Rice Solution’ mobile app with real time dashboard through AI technology.

The system has following the standard code convention, code level documentations, header of each file, algorithms, interfaces, code compression. Image sensors used through ‘Rice Solution’ mobile app for timely capturing accurate, precise and high resolution images of insect and disease of rice from different locations of the country. Then captured image or dataset are trained or supervised through AI. When any insect and disease is detected, this real time data has sent to the cloud server for further processing through an artificial neural network (ANN).

• Developed a platform as mobile app and uploaded into Google Play Store and Apple App Store (iPhone) named as ‘Rice Solution’. • Developed image analysis sensor through AI and Machine Learning Method (MLM); • Already stored and trained approximately 56000 images of several disease and insect through AI and MLM technology.

BRRI has been working relentlessly to increase food production by applying the technology of the 4th Industrial Revolution in the future. As a result, mobile phone and internet connections are reaching all over Bangladesh. Therefore, it has taken an initiative to create a mobile app based on image analysis for all types of user in Bangladesh. As a result, such initiatives will play an effective role in the food security of the country in the future.

• Engage agribusiness community for volunteer reporting; • Image collected by scientists and SA from BRRI HQ and all RS of BRRI without any incentive. So system will sustain; • Researcher - extension personnel - ICT technical expert panel has been formed for image classification and training purpose; • This technology could be the Prototype (Business Model) for other crop species of other research institutes. Other organization doesn’t required to develop same type system for avoiding duplicity; • The business model could be the named as ‘Pest e-Survilliance Platform’.