smart_society

スマート農業

農業の“工場化”に向けて,各個体の育成状況に応じた適切な環境制御・草勢管理さらには,作物の品質予測が重要である.本研究では,大規模な栽培記録データと気象データを用いた作物品質予測手法の確立および,従来目視によって把握してきた栽培状況を多地点・連続的・客観的に記録可能とする栽培指標の明確化と測定方式の確立に取り組んでいる.具代的には,農作業が有する時系列性に着目した説明変数の作成およびカスケード状モデルの導入に基づいた高精度の品質予測器構築および,圃場特有の微気象データを観測する“PASERI”を開発し,光透過率を高密度に測定することで葉面積指数との相関などを明らかにしている.これらの検討に基づき,グリーンハウスでの実証実験及びデータ解析に取り組んでいる.

To achieve an evidence-based agriculture, it is important to realize real-time monitoring of the degree of growth, abnormality detection, and quality prediction. This research aims to 1) construct a cultivation-database by using legacy farm record data sets and the AMeDAS meteorological data, 2) clarify the “index of growth” by objectively and continuously measuring the degree of growth, and 3) establish the measurement method. Specifically, we are focusing on the series characteristic of farm work and looking for methods of introducing and tuning of cascaded model, and developing the PASERI: “Photosynthetically Active Radiation Sensor for Evidence-based agRIculture” platform, which measures light transmittance over and under foliage in several greenhouses. We then model and correlate the data to leaf area index to validate the performance of our method. Now we are also developing a “Smart Greenhouse”, an application that maintains a desire environmental status in a greenhouse using macroclimate data obtained by WSNs.