AI that counts every grape berry.
An AI app born at the University of Yamanashi
Snap one photo of a cluster. AI estimates the berries you can see — and the ones you can't. Berry-thinning judgment, in anyone's hand.
No internet required. Works anywhere in the field.
Berry-thinning is a master's craft.
Read the berry count on each cluster, then remove the excess. This task — which decides a cluster's shape and quality — has always rested on years of experience and intuition. For newcomers to farming, it is anything but easy to learn.
Tsubura puts that expert judgment within reach — a tool that turns experience into a number anyone can see.
1.16M → 300k
Japan's core farming population is projected to fall from 1.16M to roughly 300k over 20 years — a structural shortage of skilled hands.
Source: MAFF (Japan)
Three steps, in the palm of your hand.
- 01
Photograph
Hold your phone up to a cluster and take one photo. No special equipment.
- 02
AI estimates
Object-detection AI estimates not only the visible berries, but the ones hidden behind the cluster.
- 03
See it instantly
The estimated count appears on the spot — an instant guide for thinning.
- Fast on-device edge AI
- Works fully offline
- Object-detection vision
- Also estimates harvest timing
How Tsubura connected a grandmother and her grandson.
21st Elementary & Junior-High Newspaper-Essay Contest · MEXT Minister's Award
A junior-high student in Ichikawamisato came across a newspaper article about counting grape berries with a smartphone. For his grandmother, who still tends her vineyard in her eighties, he downloaded Tsubura and tried it with her in the field.
The essay he wrote about it, “The Power that Connects through Newspapers,” won the MEXT Minister's Award at the newspaper-essay contest.
Technology that bridges generations. That is the kind of tool we set out to build.
“You're like a wizard. It really tells you, just like that?”
— The grandmother's words, from the award-winning essay “The Power that Connects through Newspapers” (Yamanashi Nichinichi Shimbun, November 30, 2024)
Quotation and article © The Yamanashi Nichinichi Shimbun. Excerpted here with attribution.
Where Tsubura has reached.
- Press release
- Free release of the AI berry-count app for thinning
Univ. of Yamanashi · 2024.5.30
- Television
- “Grape thinning made efficient — AI phone app released”
NHK Yamanashi · 2024.5.30 (no longer online)
- Newspaper
- “AI supports grape berry-thinning”
Yomiuri Online · 2024.6
- Newspaper
- “Grape thinning made efficient with an app”
Nikkei · 2024.6.22
- Television
- “Good news for growers — AI streamlines thinning”
YBS · 2024.5.30
- Award essay
- “The Power that Connects” — MEXT Minister’s Award
Newspaper-essay contest · 2024
From years of research to the field.
Tsubura grew out of image-recognition research at the Mao & Zhu & Buayai Laboratory, University of Yamanashi. Building on it, IoA-8Lab implements field-ready AI — and hands over the ability to operate it.
A University of Yamanashi venture
A research-origin, implementation-focused AI company.
Patented berry-detection technology
The core image-recognition / berry-detection method is registered as patent No. 7479007.
Field-validated technology
AI shaped by repeated demonstration in real vineyards.
Try Tsubura today.
A free trial — try it right away. Put an AI eye in your own vineyard.
Talk to IoA-8Lab about deployment & collaboration →