DAVAO CITY — A new artificial intelligence-powered system is set to transform the way the country's "King of Fruits" is graded and brought to the global market. The University of the Philippines-Mindanao has officially introduced AIDurian to standardize quality assessments for the booming durian industry.
The Challenge of the King of Fruits
The Philippines produces at least 80,000 metric tons of fresh durian annually, generating almost PHP 6 billion in revenue. However, choosing the best fruit has traditionally relied on manual methods, making quality grading subjective and highly dependent on the experience of individual classifiers.
Manual checking requires evaluating the moisture of the stem, the flexibility of thorns, and the sound of a dull, hollow thump when tapped. While traditional classifiers rely on these physical cues, the inconsistency of human judgment creates a critical vulnerability for the country's agricultural exports.
"These methods are subjective and difficult to replicate consistently; grading can vary among classifiers and may become a bottleneck when skilled personnel are limited," said Dr. Armacheska M. Satina, the project leader. This bottleneck often causes postharvest losses and missed export opportunities.
Introducing AIDurian Technology
Funded with a PHP 11-million budget under the DOST I-CRADLE program, AIDurian was completed in early January. The system integrates visual scanning, acoustic analysis, and deep learning to assess ripeness, surface defects, locule count, and shape without damaging the fruit.
Satina emphasized that the system provides a reliable, technology-enabled complement to manual tapping and visual inspection. By automating selected grading tasks, the technology helps support more timely sorting and marketing decisions while potentially reducing postharvest losses across the supply chain.
High Accuracy and Field Applications
The tool is integrated into two prototype platforms: INDAI, a mobile application for field-level inspection and decision support, and DuDONG, a desktop application designed for comprehensive laboratory-level analysis. This dual-platform approach ensures flexibility for various agricultural settings.
Testing results are highly promising, showing over 80 percent accuracy for maturity and acoustic ripeness. Furthermore, the system boasts 93 percent defect detection and an impressive 98 percent accuracy rate for locule analysis. These metrics validate the system's commercial viability.
Securing the Value Chain
Durian quality inaccuracies ruin the taste of the fruit and fall short of strict export standards. AIDurian addresses this by ensuring only prime fruits enter the market, directly protecting the income of farmers, processors, and cooperatives involved in the vast agricultural value chain.
By partnering with local durian farm cooperatives in the Davao Region, the project accelerates the digital transformation of the sector. It strengthens local capacity in AI applications for agriculture, paving the way toward more standardized and sustainable grading practices.
