2020 Winner

Croppie – the PhotoCropping app

Peru, Uganda

Scientific approaches to farm agronomy have greatly increased global yields, yet high data collection costs have primarily driven insights and benefits for commercial actors focused on larger-scale agricultural systems. The benefits of big data and AI remain largely inaccessible to the world’s 500 million smallholders, who typically manage significant socio-economic and climatic risks with little to no access to external data. Furthermore, COVID-19 has exacerbated the need for standardized, efficient, and reliable data collection to support yield estimates. 

With the development of the Croppie app, farmers will be able to gamify real-time crop yield data collection, encouraging smallholders, particularly youth, to generate massive datasets for AI models (PhotoCropping). Simultaneously, Croppie aims to provide clear, actionable yield predictions for smallholders, where AI picture classification is proven to provide reliable yield data.

More about the project

Scientific approaches to farm agronomy have greatly increased global yields. Yet high data collection costs have primarily driven insights from and benefits to commercial actors focussed on larger-scale agricultural systems. Smallholders rarely benefit from prediction insights. Big data and AI benefits remain largely inaccessible to the world’s 500 million smallholders, who typically manage significant socio-economic and climatic risks with no access to external data. A bottleneck is the standardized and efficient collection of reliable data to support yield estimates. COVID-19 exacerbates this.

The project will test a low-cost way to generate large volumes of ground-truthing photographic data while positioning smallholder farmers as the primary beneficiaries of the data and predictive insights generated.

The project’s app, Croppie, will gamify real-time photo data collection from smallholders by incentivizing PhotoCropping, i.e. taking and sharing photos of their crops to:

  • Generate low-cost, high volumes of real-time primary data from smallholder farmers that support AI/ML model training;
  • Cost-effectively empowering smallholders with predictive AI-generated yield insights in actionable formats, for more effective planning, improved risk management, and increased resilience.


This Inspire Challenge proposal was selected as a 2020 pilot project winner, receiving a total of US$ 100,000 to put their ideas into practice. Learn more about the Inspire Challenge Grant here.


The Alliance of Bioversity International and CIAT
Christian Bunn | Email
Anton Eitzinger
Eric Rahn
Juan Camilo Rivera Palacio

Producers Direct
Claire Rhodes | Email
Sam Webb
Brian Ngetich
Sarah Mackay
Trilce Oblitas

Luan Nio

Step by step



JAN 2021

Inclusive farmer focus groups selected for app design and testing

The team created farmer design groups to support the development and testing of the Croppie app in Peru and Uganda. Approximately 100 local farmers engaged in the design process in both countries. The focus groups were composed of 40 per cent women farmers and 50 per cent youth farmers.


FEB 2021

Smartphone photo protocols created for app

The team defined and tested the app protocols that enable farmers to take accurate pictures of coffee cherries and coffee bushes with smartphone cameras.


APRIL 2021

Primary data collection

The team will collect the primary data to support the training of the yield prediction model and verify the projections. The model will include manual coffee cherry counts, farmer logbook data, and qualitative farmer feedback data collected using the 5Q approach.


APRIL 2021

Farmer-led app design

The Alliance of Bioversity International and CIAT and IDEO.org will facilitate the farmer-led design process and develop the prototype Croppie app with the focus groups in Peru and Uganda, including the user-interface and use incentives components.

APRIL 2021

Coffee cherry and bush images used to train app’s AI

The team will collect and label a dataset of more than 1,000 smartphone photos of coffee cherries and coffee bushes.

Photo: Neil Palmer / CIAT. Coffee cherries.

APRIL - DEC 2021

Secondary data collection

Secondary climate data will be collected throughout the year from CHIRPS and WorldClim, and soil data will be gathered from the FAO, among other sources.



OCT 2020

Project awarded US$100K Inspire Challenge grant

The project was one of seven winners of the Inspire Challenge 2020 and was awarded US$100K at the fourth annual convention of the CGIAR Platform Big Data in Agriculture, 19-23 October 2020.

Gender & Youth Inclusion

  • Croppie aims to improve women coffee farmers’ access to the coffee supply chain and technology by enhancing data capture and documentation of assets, yields, and profitability of coffee farms managed by women.

  • Croppie will use data to build users yield and on-farm performance profiles that are required to access financial services. Rigorous data privacy protocols will be employed in order to safeguard participants’ data.
  • Croppie will engage committees of women farmers and youth in the design and testing processes in order to address app usability, value-add, and potential usage barriers, including potential risks associated with data sharing and user anonymization, as well as incentives to use the application and share data.
  • Croppie will also showcase farm-based and post-harvest coffee innovations led by women farmers and youth in order to empower these farmers to share knowledge and resources worldwide.