We investigated the effect of furrow bottom seeding on ridges （FBSR） on directly seeded spring-sown onion in the Tohoku region. We created two plot types sown with seeds of three selected onion cultivars–Momiji-3go, Ohotuku-222 and Maruso – sown at the bottom of furrows on ridges （Furrow plots） and at the top of the ridges （Top plots）. A furrow on ridges was 10 cm wide and 6 cm deep. Daily minimum soil temperature and soil moisture beneath the Furrow plots were higher than beneath Top plots. Daily maximum soil temperature was lower beneath the Furrow plots than the Top plots. The seedling emergence rate at 20 days after sowing and average plant height at 62 days after sowing were higher for Furrow plots than Top plots for the three cultivars. The day when at least 50% onion tops fell over （DFO） was later in the order of Momiji-3go, Maruso and Ohotuku-222 in both Furrow plots and Top plots. The DFO of Momiji-3go for Furrow plots was 8 days earlier than for Top plots, and corresponding onion bulb weights for Momiji-3go were 68.4 g and 50.5 g, showing a small difference. However, onion bulbs of Maruso and Ohotuku-222 for Furrow plots were 1.6–1.8 times heavier than for Top plots. The onion bulb shape for Furrow plots tended to be slightly more elongated than for Top plots, but this was not a problem because the length-to-diameter of bulbs was less than one. For direct seeding of spring-sown onion in Tohoku, FBSR accelerated germination and growth and increased yields more for later cultivars.

In this study, we investigated the pesticide spray operations of multicopter in hilly and mountainous areas of Hiroshima. By comparing the work efficiency of the multicopter and boom sprayer, we aim to explore the time benefits of utilizing the multicopter in hilly and mountainous areas. First, we studied how the work efficiency of multicopter and boom sprayer were affected by the size of the fields. Second, by modeling the workflow of the multicopter and calculating the time of each step in the workflow, we estimated the total time taken by the multicopter to perform the same activity in the same fields where the boom sprayer was used. By comparing the time consumptions, we had the following results: （1） work efficiency in the fields was approximately three times higher when using the multicopter compared with that of the boom sprayer. The work efficiency was lower in the smaller fields for both multicopter and boom sprayer, but the efficiency of the multicopter was relatively higher;（2） 55% less time was needed when using the multicopter compared with the boom sprayer. Based on the results of this study, pesticide spray operations by using the multicopter is more desirable in hilly and mountainous areas.

Average is often used as a representative value of work time. However, when some time values close to the abnormal value are included, the average value becomes larger or smaller than the original value. In farm work, the production process depends on conditions that are difficult to control artificially due to the influence of the natural environment. As a result, abnormal values are likely to appear in the working time, and the distribution tends to be thick on one side. In agriculture, there are several cases where it is desirable to grasp the working time, which is a typical value that can be called the original time value. Therefore, in this study, we selected a kind of mode value as a representative value of farm work time. There are many problems with the existing mode calculation methods. Usually, the mode value is obtained as a class value of the maximum frequency class after making a histogram. However, the mode value thus obtained becomes indefinite depending on how the “anchor position” is determined, even if the class width is fixed; therefore, uniqueness is not guaranteed, making it undesirable. The mean shift method and the approximate L0 norm estimation method are methods that do not use a histogram, although they also have problems. Therefore, we devised a new representative value that is a kind of mode value called the core median. The core median can be calculated using the calculation procedure at spreadsheet software worksheet in this study. In addition, we applied the method on actual farming data for its validation, and we discuss the conditions that should be taken into account while using the method.

A measuring apparatus with an external output function connected to a Bluetooth adapter and a laser distance meter with Bluetooth function can easily send measurement data to a smart device such as a smartphone or a tablet. We fabricated such an apparatus that could be easily operated by one person. For measuring plant height and seeding depth in the field, the vertical distance measuring apparatus used was a laser distance meter with a Bluetooth function fixed to a monopod and a smart device. The apparatus could measure height with an error range of about 2 mm. To measure the weight and width of an object, an electronic balance with Bluetooth adapter and an apparatus （in which the object was placed between a fixed plate and a sliding plate） measured the width with a laser distance meter and a smart device were used. The working time when measuring the weight and width of sweet potato roots using this measuring method was 58％ less than that for a visually measuring method that required using a ruler, an electronic balance and filling in a field book.