Crop loss assessment stress test.

IBISA leverages leading-edge technologies to enable scalable and efficient crop loss assessment.

IBISA’s mission is to enable agriculture insurance for agricultural entrepreneurs. We leverage technology and data to build and manage efficient, scalable and transparent parametric insurance products worldwide and in an easy manner.

One of our key values is the innovative solution for loss assessment. For loss assessment IBISA combines the use of EO products with a “crow-watching” platform. The theory behind is that errors in individual assessment caused by partial information or bias tend to be cancelled. 

The objective of this pilot project is to backtest this solution to validate it faster and widely and identify points of improvement. 


The challenge of crop insurance is to maintain loss assessment scalable, fast and cost efficient. In order to do this, IBISA is developing a “hybrid-parametric” assessment.This technology mixes automatic identification of losses using Earth Observation data, and blockchain based crowd-assessments.

On one hand, automatic assessment uses automated loss identification based on satellite images to provide an estimation of loss.This method is very fast and scalable but suffers from limitations, typically in the presence of very cloudy conditions.

In order to overcome these limitations in a scalable manner, IBISA leverages crowd-watching and is building a community of people called “watchers” to provide a second opinion on the automated assessment. This “crowd-watching” is organised with a token-economy, that rewards good watchers and penalises cheaters, incentivising all the community to behave in the interest of the farmers and the IBISA system.

Work Plan

The accuracy of assessments is based on the quality of the Earth Observation data that feeds the platform. In order to improve this, IBISA team is constantly working to improve the quality of visualisation computed based on raw data.

The second step is to compare hybrid-parametric assessment with ground truth, based on in-field assessment. This is done by comparing in-field vegetation anomaly data with the backtest of hybrid-parametric for the same location.

  • Hybrid-parametric loss assessment validation(July to October): The goal of this activity is to validate the viability of the “hybrid-parametric” or “crowd-watching” approach.
  • Earth Observation vegetation index improvement (July to October): IBISA uses multiple Earth Observation indices for the hybrid-parametric assessment. These indices are regularly improved. The goal in this activity is to define a new way of representing NDVI index, using statistical distribution of it.