AGRIDATA — Smart-Greenhouse AI Data Platform

Big-data platform for smart-greenhouse production-environment management — data curation, quality, and AI analytics

AGRIDATA built a practical AI platform for managing the production environment of smart greenhouses (Korean title: 스마트 온실 생산환경 관리용 인공지능 플랫폼 실용화), under the national multi-ministry Smart Farm package program. The goal was a data-centric, autonomous-cultivation greenhouse: collect and curate field data, learn growth models, and drive AI-based decision support and optimal environment control. The consortium was led by Kyungpook National University, with KETI and Daliworks as co-research institutions.

Platform overview: a smart-farm big-data pipeline feeds AI model management and analysis, driving a decision-support precision-agriculture platform (FES) for optimal cultivation timing and control.

Highlights

  • A full-lifecycle agricultural data platform — multi-source ingestion (self-collected, platform, and public data), storage, and automatic metadata generation
  • Eight data-quality-improvement modules plus a formal quality-metric scheme (syntactic, semantic, and range accuracy, and risk-of-inaccuracy)
  • Multi-source data alignment to reconcile differing sampling periods and types without distorting the originals
  • Cloud integration on a Naver Data Box / Connected-Server architecture, feeding decision-support and farming-execution (FES) services

KETI’s major contributions (KETI was the co-research institution for the data platform)

  • Greenhouse data-processing modules with performance evaluation, a dual-database data catalog, and quality criteria for training-data construction
  • A data quality-measurement scheme — syntactic, semantic, and range accuracy plus a risk-of-inaccuracy index — to validate data before analysis
  • Multi-source data alignment and Naver Data Box cloud integration, enabling model inference and data processing within the platform’s access constraints
KETI contribution (Year 1): greenhouse data-processing modules and performance evaluation, a dual-database data catalog, and quality criteria for training-data construction.

Funding: Ministry of Agriculture, Food and Rural Affairs / IITP
Period: 2021 – 2023
Role: Co-PI (KETI)