Please login first
Towards a Crowdsourced, Digital Coffee Atlas for Sustainable Coffee Farming
, , , , , *
1  University of Applied Sciences Mannheim, Paul-Wittsack-Straße 10, 68163 Mannheim, Germany
Academic Editor: Dirk W. Lachenmeier

https://doi.org/10.3390/ICC2024-18176 (registering DOI)
Abstract:

The presented work summarizes the results of a 15-week student project addressing the field of sustainable coffee farming. Coffee farmers often lack scientific knowledge concerning the coffee varieties they cultivate. Also, they have been growing coffee for generations, they often have limited knowledge concerning the names of their coffee varieties used on the global market. This leads to significant disadvantages in market positioning. Consequently, farmers often receive lower prices for their coffee as they cannot accurately determine its true market value. In addition, the effect of climate change forces farmers to reconsider their cultivated varieties as they can’t exhibit stable yield performance due to the changed climate. The potential quality advantages of different coffee types are not known, which prevents farmers from optimizing growing conditions specific to their climate. As part of a design thinking-based project course, a team of four students from design and computer science at Hochschule Mannheim was searching for a solution on how to overcome the forementioned disadvantages for local coffee farmers with the support of digital technology. Coffee Consulate helped the team by connecting them to farmers around the world and sharing their domain knowledge. The student team’s main idea is to bridge the forementioned knowledge gap of farmers by collecting globally distributed data about coffee species in one worldwide accessible, digital system. That way farmers could be globally connected. Their concept proposes a digital Coffee Atlas for mobile phones showing where on the planet and under which climate conditions coffee varieties are growing and how these species are named on the global market. The app allows one to identify coffee plants based on pictures uploaded on the farmer’s phone. The team developed an implementation roadmap that considered how to subsequently extend the database behind the Coffee Atlas and how to accelerate the crowdsourcing process. AI-based image recognition trained with pictures taken in a living collection of coffee cultivars, like in the botanical garden of Wilhelma (Stuttgart, Germany), and DNA sequences could serve as an initial step to pull the database off. Farmers should be motivated to upload pictures of their plants by additional services provided by the app. Thereby the information about coffee species can be crowdsourced with the help of farmers around the world. Such services could include the recognition of the plant’s health condition, as well as the estimation of the actual market price of the species based on the identification of the coffee varieties or the recommendation of species that are better adapted to the actual or expected climate. In its final implementation, the Coffee Atlas will enhance agricultural practices and economic outcomes for farmers and provide a valuable source of data to researchers around the world.

Keywords: artificial intelligence; coffee varieties; sustainable farming; digital solutions

 
 
Top