Drones: Set to make crop loss assessment smarter- A case study from Maharashtra.

STATEMENT OF PROBLEM

Due to high soil moisture retention for long period in agriculture fields, especially Soybean fields, the number of snails in farm fields have grown disproportionately high and started damaging 1-2 month old crop. Most farmers in the area have reported crop damage and compensation. Quite a large number of them have even removed the damaged crop in mid-season and were preparing for resowing some other crop. As the number of loss intimations were running into thousands, it was not possible to visit every farm and survey every farm. A dire need was thus felt if technology could help expedite the payout process.

STUDY AREA

SOLUTION ADOPTED

Drones: Vertical Take Off and Landing (VTOL) fix wing long endurance drones equipped with high resolution cameras and supported on ground with DGPS high precision location capture for accurate area calculation.

Cadastral maps: Georeferenced cadastral maps for linking farm ownership data with farms captured from drones.

Digital image processing: Aerial image mosaicking was performed using leading industry standard COTS for stitching together the images for seamless mosaic. Finally leading to computation of indices linked to crop condition.

GIS software: for overlay query and computations.
Assessment period: Survey was done in Late August, report submitted in 2-3 weeks and project concluded in November 2022.

FIELD DATA COLLECTION:

Mobile applications were used to collect representative geotagged field photographs for crop type detection and crop health assessment.

ANALYSIS OF CROP DAMAGE

This is where the unique value addition of drones is clearly seen.
Every farm had unique index value that was reflective of crop condition within its bounds. This was then aggregated to farm level crop loss %age.
The scenario of three villages turned out to be Different. While Karajgaon showed majority of farms showing up complete crop loss, the Dhoki village seemingly has very few fields with damaged crop.
Important: Fraud detection is also another angle. While genuine cases can be paid quickly, the farmers who haven’t suffered any loss but did raise compensation claim can quickly be detected and weeded out from claim assessment cycle.
We therefore, call it a solution that brings transparency into whole process of assessment and settlement.

BENEFITS OF SOLUTION

Faster: In just 3 days of time a 3 member team was able to assess 3 villages covering every farm. This otherwise would have not been possible even with a 50 member field surveyor team to visit every farm.

Independent assessment:: The Drone team is an independent assessment team deployed by State Remote Sensing Application Centre who do not align with any direct stakeholder (farmer or insurance agency).

Aerial view gives different perspective: A farm if viewed from one corner, doesn’t often give a complete perspective when compared to an aerial view.

Transparent: The results are so simplified that it can be verified by any person who is not expert including the farmer. This eliminates ambiguity in interpretation and getting into a dispute cycle.

Digital: Being all digital, it can be shared among different stakeholders easily as well as can be printed and published at will. The orthomosaics can be overlaid with Google images for better understanding of location of farm and other features on ground.

Sharable: Quick farm specific personalized one pager reports can be instantly prepared and tagged to intimations to complete assessment cycle.

Objective assessment: The main problem with visual assessment by ground surveyors is that it is subjective, and different persons might assign a different percentage of loss (Unless the crop is completely removed from ground).

Scalable: The solution is completely scalable and SOPs can be shared across with any implementation agency for consistent results anywhere. Drones are very popular now a days and easily available for deployment given the pro-active Government’s support. It can be deployed at impacted locations within 2-3 days max, howsoever remote the area may be.

Free from cloud cover impact: Though satellite datasets also can be of help here but in the event of cloud cover during Kharif season, sensors flying below cloud cover can only fetch ground information and that is possible only with drones.

SOLUTION IMPACT

Farmer gets quicker relief in case of loss. On further integration with banking data, payouts can be made directly to farmer’s account and completely eliminate middlemen and time lag.

Insurance companies are OK to pay if the losses are genuine and quantifiable.

Direct impact of the study over 3 villages has been close to 6000 but indirectly it would have benefited their families and other stakeholders that could be no less than 4-6 times the direct number.