The effect of different row spacing (35 cm and 70 cm) on the development of soybean plants was studied using two soybean varieties, Lenka and ES Mentor. By using RFID technology, which includes a smart tag with a microchip, temperature, moisture and three sensors, an antenna and a reader, we measured temperature, humidity and light in the middle of the row space of each experimental field throughout the growing season. Besides, crop development, measurements of the morphological characteristics of the plants inventory of weeds and yield weight were made. We discovered that differences between soybean varieties were more pronounced than differences between different row spacing. The application of RFID technology was proved to be accurate enough in plant growing, so together with proposed improvements RFID would allow the farmer to control agricultural land more quickly and to respond in a shorter amount of time in comparison state meteorological station.
RFID (Radio frequency identification) is an emerging technology that can have
enormous opportunities in Agri-food sector. Technological developments in the area
of networking devices, sensors and communication technology play significant role in
sustainability of Agri-food sector. RFID is one of such pervasive technology, which is
now increasingly utilized in food logistics, supply chain management, cold chain
monitoring and retail
Biosensor tags are also under development that could be used for detecting bacterial contamination of food products along the supply chain (Wentworth, 2003). The main purpose of our research was to use RFID technologies with sensors for measuring temperature, humidity in light intensity between soybean plants during the vegetation. Two varieties of soybeans ‘Lenka’ and ‘ES Mentor’ both of the same maturity class (00 were grown on 35 cm and 70 cm row spacing. We investigated the differences in plant development as well as temperature, humidity and light gradients between varieties and row spacing. Finally, we estimated the accuracy of RFID with the measurement of meteorological conditions on nearby meteorological station. 2
The soybean experiment was carried out on the field owned by a local farmer, it is located in the municipality of Hoče-Slivnica (Slovenia). The plot ‘Štuk’ (46o30’28’’N, 15o039’19’’E) has 2.37 ha in area with the average altitude of 272 m and the average slope is 1% and 0.6 °, respectively.
For the purpose of the experiment, the south-eastern part of the parcel was selected and divided into four experimental sub-parcels with 1200 m2 each (Fig. 1). The fields were visibly divided by 0.75 m wide "border bands", where no crop was presented. The relatively large area of the experimental fields allowed for a more precise choice of location for the placement of the measuring devices and represented a more representative sample for the analysis of the results.
The prerequisite of soybeans was winter wheat, followed by white mustard after harvest. The experimental boxes were separated by wooden stakes, marked with variety and row spacing.
We used a measuring system consisting of a semi-passive smart tag with an integrated circuit, in our case it was a SL900A microchip, which has an integrated temperature sensor and allows the connection of additional external sensors. Furthermore, the measuring system consists of sensors, antenna and battery. We collectively call this set a smart badge or Tag (Fig. 2). The measuring system also includes an antenna reader, in our case, the DRM-900A, which transfers data from a badge to a computer. The smart badge was equipped with a humidity sensor (EMD4000) and an illumination sensor (TEMD6200FX01). The described RFID measuring system automatically measures and stores data every 2 hours. In this way, the memory was sufficient for 14 days of data storage. Battery life is up to five years, depending on weather conditions and measurement time interval. The data transmission range can be up to 100 m with a properly performing antenna, and the transmission starts when the radio frequency signal of the reader is received. The system operates at ultra-high frequencies (UHF 860-960 MHz) according to the master-slave principle.
The sensor is a semiconductor element - a closed-polarized diode that converts light into electrical current. Sensor size is 2mm x 1.25mm x 0.85mm. The spectral sensitivity of the diode is adapted to the response of the human eye. EMD-4000 Moisture Sensor The sensor substance is a special polymer deposited on a bismuth coated ceramic substrate. The polymer changes the impedance in the presence of moisture. The smart badge excites an alternating current sensor and measures impedance as a function of relative humidity. The size of the sensor is 10 mm x 5 mm x 1.5 mm.
As seen from Fig. 4 the highest average plant height was measured by ‘Lenka’ 35 cm (110.1 cm). The highest number of branches per plant was recorded with ‘EC Mentor’ 35 cm (57). The maximum distance of the first branch from the ground (13.5 cm) was measured by both varieties at a distance of 35 cm between the rows. The ‘Lenka’ variety had more shoots that are lateral at both interspace distances than the ‘ES Mentor’ variety. variety shows a bigger coverage of the row space at a distance of 70 cm. Despite the average number of buds per plant of the ‘Lenka’ variety, this covered less row space. ‘Lenka’ 70 cm covered an average of 59.6% of row space, while the ’ES Mentor’ 70 cm averaged 70.7% of row space, respectively.
Table 3 shows the average values of humidity for the individual sensors at each three measurement intervals. The data measured correlate opposed as the temperature development due to the differences in coverage between varieties and row spacing. Thus, the highest average relative humidity was measured in ’ES Mentor’ 35 cm but did significantly differ only from ‘Lenka’ 35 cm as well as from the data of meteorological station. Later shows smaller values due to the open meadow in which the data very captured.
As seen from Table 4 the light intensity is significantly lower below the leaves in the zone of sensors in all varieties and row spacing. Additionally, from the first to the last measuring period the development of leaves decreased the light intensity significantly in the same treatment. However, in the second and third period the light intensity was significantly lower in ‘ES Mentor’ comparing to ‘Lenka’ in both row spacing. This corresponds also with the higher number of pods in ‘ES Mentor’ (Table 1).
On October 13 the soybean was harvested and weighed directly on the field. It contained 12.5 % of moisture so later it had to be additionally dried out to storage moisture of 9 %. From Table 5, showing the yield of soybean with 9 % moisture, we can see that the ‘ES Mentor’ had significantly higher yield than ‘Lenka’ in both row spacing. On the other hand, there was no difference between the row spacing among the same variety, either.
The effect of different row spacing of two soybean varieties on development of plants was studied using RFID technology to measure temperature, light and humidity in the middle of row spacing. Since the growing phase R1, we estimated the height of the plants, the distance of the lower pods from the ground, the number of lateral buds and the number of pods per plant. We found that the row spacing did not have a significant effect on the development of the crop within the same variety. However, the biggest differences were observed between the two soybean varieties.
RFID showed to be a suitable tool for accurate measurements of temperature and humidity more accurately than the official meteorological station, which might serve for predicting the start of many fungal diseases on time.
The UHF-RFID technologies have also some disadvantages. Above all, the problem is the reliability of the connection, since electrically conductive materials represent a barrier to electromagnetic waves (reflection and wave absorption). Thus, moisture in plants is also a barrier. These problems do not have a Magnetic Assembly (HF) based badge.
The second problem represents the storage capacity of RFID as well as the reading and transferring of data every few weeks. The easiest way would be to read data from RFID in the field using a drone with a built-in UHF reader. The drone would be connected to a home computer via a WiFi network, which would determine, for example, the time, method or means of spraying, based on the built-in algorithms and data collected. If such a drone were equipped with hyperspectral imaging cameras, then algorithms could be installed to separate and detect the state of plants based on spectral signatures. This would give a reliable prediction of the necessary measures, which would be an effective contribution to the future vision of technologically advanced farms and smart villages.
Acknowledgement. This article was partly created as a result of the applied project ‘Reducing of draught stress and increasing of soil fertility by introducing conservation (conservation) soil tillage into sustainable agriculture (V4-1815)’ funded by the Public Research Agency of the Republic of Slovenia and the Ministry of Agriculture. Forestry and Food of the Republic of Slovenia. Authors also thank to Mišel Pleteršek for his vital contribution in managing practical experiment in his field.