Kisan Suvidha is an automated, transparent and low-cost system for farmers in India that can ensure timely crop insurance payments by using AI and blockchain, developed by a team of two women techies.
Recently at an AI and blockchain hackathon organised by Icertis, two women techies under Team Heuristic showcased their unique solution aimed at helping millions of Indian farmers.
Laisha Wadhwa, A Microsoft Student Partner and Garima Yadav, Graduate Analyst at Barclays came together for an all-women team (Team Heuristic) for a hackathon organised by Icertis. The duo built Kisan Suvidha — an automated, transparent and low-cost system that can ensure timely crop insurance payments by using AI and blockchain. The solution enables insurance policy managers to process claims (manually raised by farmers) based on its weather risk predictions.
Using Kisan Suvidha, the farmer just needs to upload its crop insurance policy and a picture of his farm. The app automatically takes in the location of the farm from the policy and notes down data such as weather condition, rain, humidity and temperature of a location for the next seven days and determines the severity on a scale of 1 to 5 using AI.
The thresholds for payouts are set accordingly to the severity levels, and the payments will be triggered using blockchain smart contracts. An important step of the proposed solution was: calculating the percentage amount for payment from the insurance amount based on weather risk predictions.
Kisan Suvidha Leverages Advanced Analytics On Different Data Sets
Kisan Suvidha uses multimodal weather data — weather patterns like rain, temperature, humidity, soil moisture, wind pattern and takes all these data points across a seven day advance period. The system also comes with features like, farm condition monitor (weather and soil), timely push notifications regarding weather and an easy verification system that authenticates the farmer’s identity using five different kinds of identification proofs.
The data was obtained from APIs like open weather API and HERE API. We also use robust computer vision classification models to identify the identity proof of the farmer to authenticate a valid user of the Kisan Suvidha system. The classification model works on five different types of identification proofs: Aadhar Card, PAN Card, Voter ID card, Driving License and passbook. It then uses the Google Cloud vision API (OCR: Optical Character Recognition) to automatically fetch all the important details from the Crop Insurance policy.
“Two very interesting aspects of our solution are – the facility for manual claim from the farmer in case the data from the weather APIs falters. Secondly, post-filing a claim the farmer can see the status of all his claims under the Insurance claim history section,” said Team Heuristic.
Using Blockchain Smart Contracts
After the details are fetched, they are stored in blockchain distributed ledger, once the farmer details are a part of the blockchain, they can’t be altered, hence the data is tamper-proof and secure. Kisan Suvidha also makes use of smart contracts, which automatically get triggered when certain conditions are met, in their case, when climatic conditions cross a threshold value.
Team Heuristic members told, “Our solution makes use of smart contracts, which automatically gets triggered when certain conditions are met, in our case, when climatic conditions cross a threshold value.”
Using the data for the next seven days the probability of risk of extreme weather conditions is predicted using the multimodal fusion of all-weather data for AI. Based on the severity of the risk prediction, the per cent of insurance amount (ranging from 45% to 100%) is disbursed in the wallet of the farmer through a blockchain based smart contract, which allows for credible transactions without any third-party intervention. Hence there’s no need to pay intermediaries (middlemen), and it saves the farmer’s time and long waits at government offices in India for money.
Helping Farmers Keep Track With Blockchain, AI & A Web App
Kisan Suvidha not only helps farmers maintain multiple crop insurance policy but also helps them understand the weather patterns in their area in an easily comprehensible dashboard. The farmer also has the privilege of monitoring his farm (logging historical data points – like temperature) by selecting his farm on a simple to use map integrated using the Dashboard for Agricultural Monitoring available in Agro API from agromonitoring.com.
“We are planning to expand our solution by embedding a buddy, where farmers belonging to the same area with similar soil type and crops will be grouped together so that they can discuss irrigation frequency and other solutions to avoid crop failure,” told Laisha and Garima from Team Heuristic.
About The Duo That Makes Team Heuristic
Laisha Wadhwa is a machine learning and deep learning aficionado with an interest in leveraging the power of algorithms and mathematics to solve real-world problems. Currently, she’s an undergraduate at IIIT Sri City, and works with big-data analytical tools, creates and maintain models, and explore new data sets to hypothesize new solutions.
Garima Yadav, on the other hand, is an experienced Lead Web Developer with a demonstrated history of working in the computer software industry. According to Garima, she is skilled in Java, SQL, Spring Boot, IBM DB2, PHP, and C++.