DCON2025 Finals Appearance — AgriNode

We competed in the finals of the 7th National Institute of Technology Deep Learning Contest (DCON2025) as "Rebounder," a joint team from Gifu National College of Technology and Fukushima National College of Technology.

About "AgriNode"

AgriNode is a system designed to solve the problems of theft and underpayment at unmanned farm stands.

Background and Challenges

Unmanned farm stands are an important sales channel for aging agricultural producers looking to expand their reach. However, issues such as non-payment and product theft occur frequently, resulting in serious revenue losses. Conventional surveillance cameras only allow for after-the-fact review and are unable to detect fraud in real time.

Solution

AgriNode combines AI cameras with load cells (weight sensors) to deliver the following capabilities:

• Product removal detection: AI cameras accurately detect when a product is picked up
• Payment monitoring: Load cells detect coin insertion and cross-reference it with product removal
• Real-time fraud detection: Instantly detects underpayment or product theft and sends smartphone notifications
• Low cost, easy installation: Can be retrofitted to existing unmanned farm stands

Smartphone Integration

Sellers can monitor sales in real time through a smartphone app, enabling remote sales management from anywhere.

Mentor

Yosuke Okada, Representative Director and CEO of ABEJA, Inc. (株式会社ABEJA), served as the mentor for this project.

About the Team

The team name "Rebounder" is the same as our company name. This challenge at DCON became the origin of Rebounder, Inc. We will continue to apply the technical skills and entrepreneurial spirit cultivated at our colleges of technology to solving societal challenges.