UP Mindanao AI Tool Grades Durian Without Cutting It

By admin ·

Picking a good durian has long depended on a trained eye and a practiced knock. Now, researchers from the University of the Philippines Mindanao (UPMin) have built an artificial intelligence tool that can help do the job faster and more consistently, without cutting the fruit open.

The system, called the AI-based Non-Invasive Grading System for Durian, or AIDurian, combines computer-operated visual scanning, acoustic analysis and deep learning. It checks key fruit attributes such as ripeness, surface defects, locule count and fruit shape, all without damaging the durian.

The Philippines produces at least 80,000 metric tons of fresh durian every year, with production worth almost ₱6 billion, serving both local and global markets. Yet grading the fruit still relies mostly on manual methods. Classifiers look for a moist stem, pliable thorns and distinct bulges, and tap the fruit to listen for a dull, hollow thump that signals ripe flesh separated from the shell.

“These methods are subjective and difficult to replicate consistently; grading can vary among classifiers and may become a bottleneck when skilled personnel are limited, particularly during peak handling periods,” said Dr. Armacheska M. Satina, the project leader.

AIDurian runs on two prototype platforms. The first, Inspection of Durian Using AI (INDAI), is a mobile application for field-level inspection and decision support. The second, Durian Desktop-Oriented Non-Invasive Grading System (DuDONG), is a desktop application for more detailed analysis.

In testing, the technology was reported to reach more than 80 percent accuracy for maturity classification and acoustic ripeness, 93 percent for defect detection and almost 98 percent for locule analysis.

Accurate grading matters because errors can affect the taste and ripening of the fruit and cause it to fall short of export-quality standards. This can lead to losses and lower incomes for durian farmers, processors, cooperatives and others along the value chain.

“It provides a technology-enabled complement or alternative to traditional manual tapping and visual inspection, helping improve the speed and consistency of grading durian,” Dr. Satina said. She added that by automating selected grading tasks, AIDurian can support more timely sorting, handling and marketing decisions while potentially reducing postharvest losses.

The project was implemented by UPMin under the Department of Science and Technology’s (DOST) I-CRADLE program with an ₱11-million budget and was completed in early January this year. The DOST-Philippine Council for Agriculture, Aquatic, and Natural Resources Research and Development monitored its implementation, and the team worked with local durian farm cooperatives and other stakeholders in the Davao Region.

“AIDurian supports the digital transformation of the durian industry, strengthens local capacity in AI applications for agriculture, and provides a potential pathway toward more standardized and sustainable grading practices,” Dr. Satina said.

AIDurian is among the highlights of this year’s National Science, Technology, and Innovation Week, which DOST will hold on November 25 to 29 in General Santos City. The event showcases homegrown technologies and research that bring science-based solutions closer to Filipinos.