Buba, an AI-powered platform, was developed during the International Potato Center's (CIP) AI-griculture Challenge Hackathon in 2024 and announced as the winner in the Crop Improvement Science Goal category on May 23, 2025. The tool addresses bottlenecks in potato breeding by using artificial intelligence to support the selection and optimization of varieties.
The hackathon attracted over 60 teams worldwide and focused on AI applications for food security. Buba was recognized for its potential to accelerate the development of resilient potato varieties under climate and disease pressures. The team received US$4,000 in prize money and opportunities for scaling through CIP's network.
Potato breeding typically requires long timelines, with manual trait evaluation, cross-breeding, and multi-season field trials. Buba integrates predictive modeling and data insights to shorten cycles and improve targeting of crosses.
Buba's system applies machine learning to potato genetic data, environmental factors, and breeding outcomes. It includes a trait optimization engine to simulate cross combinations, climate adaptation modeling using weather projections, and disease resistance prediction for pathogens such as late blight and common scab. The platform can also run predictive performance simulations for specific agroecological zones.
The interface is designed with dashboard functions and potential API integration with CIP databases, including potato and sweetpotato catalogues. This could make the tool accessible to breeders in low-resource environments.
According to CIP, traditional breeding can take 10–15 years for a new release, but targeted predictions could cut this by 30–50%. The platform also aligns with climate and disease resilience goals, potentially supporting varieties with drought tolerance or reduced fungicide requirements.
Data completion is another focus, with natural language processing (NLP) used to fill gaps in trait datasets. The prototype is designed for scalability, allowing local data inputs for customization in different regions, including East Africa and the Andes.
Buba was developed by Shadrack Odikara and Meshak Emakunat, working with CIP mentors and judges. It reflects the hackathon's emphasis on germplasm dataset completion and predictive modeling for yield and nutritional quality. CIP has indicated possible integration into its Genebank request system or use in field trials.
Future development could expand Buba with tools such as drone imagery and genomic sequencing. CIP expects pilots from 2026, intending to contribute to potato variety development for the 1.5 billion people who rely on the crop.
Source: Potato News Today