Satsuma mandarin cultivated in Nanki, Mie Prefecture, Japan, is a delicious and highly profitable crop. However, the lack of new agricultural workers makes it difficult to preserve specialized agricultural skills and transfer them to new generations of workers. To support farmers, we introduced a small, stationary agricultural robot, "Field Server", which measures environmental and growing conditions in real time in orchards that used a mulch and drip irrigation system able to produce high-quality fruit despite variable weather. We developed an important cultivation diagnosis tool to support farmers based on the hardness of fruit, and monitored soil and tree water stress by means of time-domain reflectometry, (TDR), evapotranspiration, and soil moisture stress, as well as meteorological conditions. Using these parameters, we developed a water stress index for satsuma mandarin. To provide practical diagnostic techniques to support orange cultivation based on the measured data, we developed a network of sensors for measuring soil moisture content and tree water stress. We also investigated the relationship between fruit quality and the soil and tree water status to test their effectiveness as a diagnostic index. We also compared a convenient soil moisture sensor (Watermark) based on TDR and showed farmers how to install the sensor in the orchards. It was possible to continuously measure the soil moisture content except where multiple water supplies complicated the measurement process. We also found that tree water stress could be accurately estimated using TDR to measure the soil and tree moisture contents.
