Artificial intelligence (AI) is entering India’s arecanut supply chain, with an Indian startup introducing an optical sorting system designed to improve quality inspection, processing efficiency and automation. Agrograde, which develops AI and physical automation technologies for agricultural supply chains, has installed its first commercial arecanut sorting system at Vikas Traders in Mangalore, Karnataka.
The system can process approximately one tonne of arecanut per hour, enabling the Mangalore facility to process close to eight tonnes a day. It uses computer vision and AI to inspect individual nuts, identify defects and automatically separate rejected material from accepted produce.
The technology targets a longstanding challenge in the arecanut trade: maintaining consistent quality while processing large volumes. Unlike conventional colour-sorting machines, which primarily identify differences in colour, Agrograde’s system is designed to detect external defects that may not be distinguishable through colour alone.
How AI Powers the Sorting Process
The system uses Agrograde’s patent-pending Vector technology to capture 10–12 images of each nut. Proprietary AI models analyse these images within milliseconds, identify visible defects and guide the mechanism that separates rejected nuts from the accepted lot.
The models have been trained using hundreds of thousands of crop images and run on high-performance computing infrastructure powered by NVIDIA.
Rakeshkumar Barai, Co-Founder of Agrograde, said the company studies crop characteristics, including how produce rolls and glides, variations in size and shape, dust levels, and the force required for gentle sorting. He said developing the arecanut sorter took almost five months and three product iterations, with research focused on existing quality-assessment practices and the crop’s specific requirements. “This AI sorter has made warehouse operations almost completely autonomous,” he said.
The machine is based on Agrograde’s Vector architecture, which ensures high surface coverage as an object moves through the vision system. It is also designed to handle considerable variations in the size and shape of produce.
Aniket Ghadage, Head of Product Engineering at Agrograde, said the Vector architecture follows a design philosophy that makes sophisticated machines easier to operate and maintain. It is also designed to function in challenging environmental conditions, including heat, dust and humidity.
Accuracy Improves with Operational Data
Agrograde developed the system in collaboration with Vikas Traders, a Mangalore-based agribusiness with decades of experience in the arecanut trade. The partnership helped adapt the technology to actual warehouse conditions, including dust, foreign matter and high humidity.
When initially commissioned, the system recorded approximately 75% sorting accuracy. Following updates to its AI models using commercial operating data, accuracy increased to around 89%. The company expects further improvements as additional data becomes available.
Before automation, Vikas Traders depended heavily on manual inspection. According to the company, one worker could sort roughly two bags a day, while a team of around 15 people processed approximately 40–45 bags daily. The availability of skilled labour also affected inspection, packing and dispatch schedules.
The automated system provides more predictable processing capacity and generates a quality report for every lot, covering size distribution, identified defects, defect intensity and the number of nuts packed.
Raj Padaliya of Vikas Traders said, “I was curious to see if AI and robotics could be deployed in the arecanut industry. I was looking for technologies that could make our entire warehouse autonomous and fast. We were excited to learn that there was an Indian startup working on this, and we worked closely with them to try out multiple versions. Finally, we succeeded, and now our sorting process is almost completely autonomous.”
Expanding Agricultural Automation
Agrograde develops AI, computer vision and automation solutions for post-harvest operations involving onions, potatoes and apples. The company recently deployed its 150th machine across 15 Indian states.
The arecanut installation expands its work into a crop where quality assessment has traditionally relied on manual inspection. It also illustrates how AI applications in agriculture can extend beyond cultivation to sorting, grading and other post-harvest activities.
For traders, the technology combines processing capacity with more systematic quality monitoring. Its wider impact will depend on its performance across different lots and operating conditions as adoption expands.