KRESIP — Korea–Spain Smart-Farm AI Platform

AI-based precision-agriculture platform jointly developed by Korea and Spain

KRESIP developed an AI-based smart-farm precision-agriculture platform through a Korea–Spain international joint R&D collaboration (Korean project title: AI 기반 스마트팜 정밀농업 플랫폼 기술개발). The platform spans the full pipeline — field data collection, deep-learning analysis, and cloud–edge service deployment — validated on real farms in both countries.

Platform architecture: field data collection → deep-learning analysis and model deployment → cloud–edge service for smart-farm decision support.

Highlights

  • World-first flower-cluster (화방) tracking model for strawberries, plus a fruit-detection model — contributing to an international standard
  • A cloud–edge platform on a Kubernetes / AWS EKS microservice architecture for deploying and managing smart-farm AI services
  • A field rail-camera image-acquisition system feeding the growth-recognition models
  • Dual testbeds in Korea (strawberry) and Spain (vineyard), with a series of commercialized products (2021–2024)

Consortium

  • Korea: KETI (lead), Sejong University, Daliworks, Naretrend, and Kyungpook National University (consigned)
  • Spain: CT Ingenieros, Bodegas Bohórquez, and the University of Salamanca
Korea–Spain collaboration structure: shared use-cases and work packages across the two consortia, with jointly developed field hardware.

What we built

  • A cloud–edge edge-computing platform with a microservice architecture (Kubernetes / AWS EKS) for smart-farm AI services
  • Agricultural growth-recognition deep-learning models and a rail-camera field imaging system
  • A world-first flower-cluster tracking model and a fruit-detection model, contributing to an international standard
  • A time-series agricultural data platform supporting 10+ time-series processing methods, with analysis and visualization services
  • A water-stress-index complex environment-control system
World-first flower-cluster (화방) recognition and tracking: object detection, cluster identification, and tracking on the strawberry testbed (KETI).
Edge-resource optimization for running the AI services on constrained field hardware (KETI).
PoC and testbed operation: Kubernetes-based deployment of the AI services in the field (KETI).

Testbeds and outcomes: the platform was validated on a strawberry farm in Korea and a vineyard/winery in Spain. The project produced world-first technologies and international standardization, domestic and international publications, and a series of commercialized products (2021–2024). Results were promoted at IFA Berlin 2022 and KES 2023 among other venues, and seeded follow-on Korea–Spain / Korea–EU proposals (e.g. Horizon Europe).

Funding: KIAT
Period: 2021 – 2024
Role: Korean PI (KETI)