Data-driven solutions for high-tech agriculture

Novi Sad, Serbia
Established in 2019

cropt.ag

Teaser of the company

Humanity is in need of innovative solutions to provide healthy food for its growing population. Cropt was founded with a clear vision to change agriculture, one of the most conservative and least digitised areas of industry. It believes that using big data and state-of-the-art machine learning algorithms, agriculture can become greener and more profitable at the same time, for the welfare of farmers, companies, our society and our planet.

Cropt has developed the following blockchain-based prototype in the scope of BlockStart:

ALLIANCE – AgricuLturaL InsurANCE

alliance.cropt.ag

ALLIANCE was formed as an acronym of AgricuLturaL InsurANCE, but it is much more than that. It is a product that uses data and blockchain to ALLY the often conflicted sides, farmers and insurance companies.

ALLIANCE is a data-driven plug-and-play crop insurance product that automatically detects drought on the field and issues a payout through smart contracts, thus a) eliminating the need for large teams of loss adjusters and large numbers of on-site visits for claim verification b) cutting the operational and administrative costs for the insurance company and ensuring objectivity and transparency through designing more accurate and personalized contracts c) securing trust between farmers and the insurance company, which is the most hindering factor for the uptake of crop insurance.

The system is based on blockchain technology, remote sensing and big data analytics. Namely, processed climate and satellite data indicate the existence of drought-induced damages in maize and the percentage of the damage is calculated. Based on the machine-learning-driven yield prediction model, the difference between the actual yield and the yield that would have been achieved if there had not been any damages, is calculated and this difference is paid out to the farmer, through blockchain-enabled smart contracts. This system was developed as a user-friendly, crop insurance product and deployed to alliance.crop.ag. It is envisioned as a plug-and-play product, where the only integration that will be required is that with the company’s client database.

Technical development during Prototype stage

The technology developed within BlockStart is a prototype for a plug-and-play product for digital crop insurance. The prototype focuses on one peril, one region and one crop (drought/Serbia/maize). With a planted area of ~1.2Mha, maize is by far the most significant crop, with the country being in the top 10 list of global maize exporters. In the next phase of development, the system will be scaled to other perils, other regions and other crops. The year that was analysed is 2017, the last year in which drought was present.

Drought detection was based on two sets of parameters:

  • Climate triggers: low precipitation and high temperature are the initial triggers for considering the drought in the region.
  • Satellites triggers: Low NDWI (Normalised Difference Water Index) and EVI (Enhanced Vegetation Index) values. NDWI was developed with this particular purpose, to detect the absence of water in a field (threshold <0.2), while EVI is a general vegetation index that reflects the overall crop status (threshold <0.4).

Based on the satellite thresholds, segmentation of the satellite images was performed and regions that suffered drought stress were detected.

Cropt’s plan for the next 12 months is to incorporate other sources of risk in the system and to do a pan-European scale-up, where the product could be used anywhere in the EU without the need for additional calibrations.

The company already had negotiations with several agricultural and insurance companies and it plans to test the system with both its clients (insurance companies) and the final end-users (farmers). This will validate both the business model and the technical approach.

Business development during Prototype stage

  • The business model and the potential for adoption were thoroughly analysed, and a series of talks with potential clients (insurance companies) and end-users of the technology (farmers) were held. This was extremely helpful in defining the business model, as well as fine-tuning the product-market fit. Talks included other companies as well, for side business models and project scalability.

    The primary business model is generating the revenues through franchise and directly targets the insurance companies. Our innovation will come in the form of a software product that can be integrated into the insurance company’s IT system. The two following business models were drawn in collaboration with an interviewed stakeholder, a C-level representative of an insurance company, who wished to remain anonymous.

    Firstly, the product will be offered to farmers via insurance companies, which will deal with customer acquisition and revenue collection. A percentage of the revenue (e.g. 10%) will be diverted to Cropt as the franchise fee and for the system maintenance. The secondary business model targets the insurance companies indirectly, via reinsurers. Insurance companies typically transfer a certain amount of the premiums (e.g. 80%) to reinsurance companies, which collect premiums from a number of companies, thus diversifying their portfolio and lowering the risk of the system as a whole. It is in their interest to include as many insurance companies in their system as possible, and collect as much premium funding as they can. For this reason, they are offering new product lines to insurance companies, which are expected to acquire new customers and increase their own earnings, thus subsequently increasing the earnings of the reinsurance company.

  • Cropt had 11 meetings in total, with:

    1. Insurance companies. This is the company’s primary group of clients and the discussions were very fruitful, as they helped Cropt to better understand the problems with crop insurance they are facing in their daily activities
    2. These meetings helped the company understand the problem from the conflicting point of view and get to know what kind of insurance schemes would satisfy the end-users
    3. IT companies. Cropt spoke with the fellow IT companies that are also engaged in blockchain driven insurance in other sectors than agri-food, to get acquainted with the obstacles they faced and the lessons learnt
    4. Other potential adopters. With them, the company developed alternative business strategies for secondary business models that would support the primary one and generate additional revenues.
    5. Investors – for taking the startup to a higher level.
  • For the Alliance project, Cropt hired 2 additional people, one data scientist, who led the development of the drought index and satellite processing pipeline and 1 developer, who led the development of Alliance platform and integrated blockchain, satellite image processing and other services into a single system. The company’s plan in the next 6 months is to hire an additional business expert, who would help Cropt present its product adequately, ensure traction and acquire additional investors and clients.

Meet the Team

Oskar Marko
Oskar MarkoCEO
Oskar Marko, PhD (co-founder, CEO) has a background in data science. He developed the system for optimisation of farmers’ sowing strategies that is the Cropt’s core technology and that won the 1st prize at the Syngenta Crop Challenge
Sanja Brdar
Sanja BrdarCTO
Sanja Brdar, PhD (co-founder, CTO) has a background in big data analytics. Her algorithms based on data fusion and deep learning helped develop a variety of systems from different fields, ranging from HIV estimation based on mobile data analytics to yield prediction based on satellite image processing
Milan Ignjić
Milan IgnjićData Analyst
Milan Ignjić (data analyst) is an expert in the field of data science. He graduated from University of Novi Sad as the best student of his generation and his work in Cropt focuses on evolutionary optimisation and image processing based on deep learning
Igor Trpovski
Igor TrpovskiSoftware Developer
Igor Trpovski (software developer) is an expert in the field of data-driven software development. After gaining industrial experience after his studies of electrical engineering at University of Novi Sad, he joined Cropt, where he is developing the system for yield prediction based on climate, soil and field data

Cropt’s achievements

For more information, download the company’s portfolio:

BlockStart allowed us to step into the world of blockchain-driven crop insurance, a futuristic concept that we always wanted to exploit, but never had the resources and appropriate support

Oskar Marko, CEO, Cropt

Total funding received by Cropt d.o.o. under BlockStart: €16,000 (€1,000 for Ideation Kick-off and €15,000 for Prototype stage)