Coffee cherry ripeness at a Costa Rican farm with Hungarian roots

After hand-picking, a basket of coffee cherries may look ready for processing while holding fruit at several stages of maturity. At Bányai Coffee in Costa Rica, the team wanted a clearer way to assess coffee cherry ripeness in harvested lots. Could a smartphone photograph make coffee cherry ripeness easier to assess and record?

A Costa Rican farm with Hungarian roots

At the BA.CO.RE research farm in Lourdes de Naranjo, the Bányai team manages experimental microplots. The estates range from 1,410 to 1,550 metres in elevation. Workers pick the cherries by hand, and the processing team needs to judge the quality of each incoming lot. Cserta wanted pickers and buyers to have a clearer basis for discussing the mix of harvested fruit.

Two lines of field experimentation came together on the estate. A living-algae foliar treatment targeted coffee plant development, while AGRI-FOOD.AI and coffee specialists developed Maduro, a mobile application for visually assessing harvested cherries and tracking coffee cherry maturity.

What did the algae experiment show?

Cserta observed that newly planted coffee shrubs sprayed with a living-algae foliar product from Fermentia Laboratories grew taller and appeared stronger than expected. Without a matched untreated control group, quantified growth logs or yield comparisons from those plots, that visual difference remains a field observation rather than a controlled trial result.

Published trials on algae biostimulants in coffee show why controlled comparisons matter. A 2021 Coffee Science study of coffee seedlings reported stronger shoot and root development after applying a mixed foliar fertilizer and algae treatment for 94 days. By contrast, a 2026 field study of Arabica coffee using a different freshwater-algae preparation found no statistically significant gain in plant growth or yield. Because neither experiment tested the specific Bányai formulation, product composition, dose and local growing conditions must be evaluated before drawing conclusions.

Why does coffee cherry ripeness matter?

After harvest, an incoming lot can contain green, partly ripe, ripe and overripe cherries. Knowing their proportions helps growers and processors describe the batch and decide where closer coffee cherry sorting is needed before pulping or drying. Uneven maturity can affect fermentation and the consistency of the finished lot.

Skin colour offers a practical visual cue, whether ripe cherries turn red or yellow in a given variety. Still, a 2024 Plants study of Arabica cherry sugars found that soluble sugar changes in the fruit skin did not mirror sucrose levels inside the bean. A deeper red skin colour is therefore no direct measurement of bean sucrose concentration or final cup score.

Physical sorting and representative sampling remain essential alongside visual checks. The International Trade Centre Coffee Guide notes that water flotation separates lighter floaters from denser sinkers, although unripe green cherries often sink with ripe fruit and still require screening.

How Maduro uses machine vision for coffee in harvest decisions

Developed by AGRI-FOOD.AI with coffee specialists, the Maduro app applies machine vision to photographs of freshly harvested coffee cherries. The current Maduro product description specifies five ripeness categories and digital records that users can save, organise and export, with offline use supported in remote plantation areas.

This image-based ripeness assessment describes the visible surface of the photographed sample. It does not measure internal bean sugar content, hidden insect damage or cupping quality. Ambient lighting, overlapping fruit, sample selection and cultivar differences all influence what the phone camera captures, so any purchasing or lot-acceptance workflow requires site-specific checks against a manual reference.

On the plantation, Cserta described introducing photo-based checks to the picking crew and proposing that maturity results inform conversations about quality and payments. Independent accuracy data for these conditions are unavailable.

From a harvest sample to a coffee cherry ripeness record

Begin with a clearly identified lot and a representative sample taken from the harvested batch. Spread the cherries on a flat surface so individual fruit remains visible, use even lighting and avoid strong glare on the fruit skin. Take the photograph, review Maduro’s five-category output and save the digital record with the lot details. Compare that distribution with manual grading of the same sample across varieties and harvest days before linking the numbers to picker payments or lot acceptance.

The people behind the project share more details in the AGRI-FOOD.AI coffee farming podcast, featuring Péter Cserta and Gergő Boromisza. To review the mobile workflow and trial options, visit the Maduro product page.