Auto Spray and Variable Rate Prescription Tests
James Pottage, Provincial Airways
Doug Mackay, Bayer CropScience Inc.
Lai Nguyen, AG-NAV Inc.
The practice of variable rate treatments (VRT) has been around for some time and applied in ground precision farming. However the adoption of VRT is still limited for aerial agricultural applications. VRT can apply to seeds, crop protection (CP), or fertilizer. Farmers normally apply fertilizers based on the average yield of a whole field. However, the land may be more or less productive in different locations on the same field depending on soil and moisture conditions. When supplying crop protectants to farmers, Bayer requests they be applied in the most effective and responsible manner according to label directions.
At the Canadian Aerial Applicators Association convention in February 2017, Doug Mackay, an agricultural engineer with Bayer’s Digital Farming department, met with Lai Nguyen, an engineer of AgNav and they discussed testing Bayer’s prescription files in AG-NAV’s Platinum, the most advanced GPS guidance system developed by AG-NAV for aerial application. After receiving a sample file from Mackay, Nguyen tested it in the Platinum and found it working well. However the question was how they could verify the performance in-field. After the convention, Nguyen introduced James Pottage, the owner of Provincial Airways in Saskatchewan, Canada to Mackay. Pottage is also a young and experienced pilot who is always advancing technology to find a better way to serve his customers. After Mackay talking about the purpose of the prescription test, Pottage accepted the invitation to join the team. Provincial Airways has airplanes equipped with a Platinum system, auto spray system and flow controller. That’s all we needed for a prescription test.
James Pottage of Provincial Airways in Saskatchewan Canada.
The first test was intended to see if a prescription file made for ground precision farming could be used for aerial application. The prescription file defined the spray and no-spray polygons as shown in Figure 1.
The spray polygons were prescribed with the rate of 4.0 gallons per acre (GPA) and no-spray polygons 0.0 GPA. There were many small polygons on the VRT map. With the aircraft flying at 150 MPH or 240 km/h, it took only half-a-second or less to cross the variable rate boundaries.
Would the auto on/off spray respond fast enough? How about the applied rate? With continuous changes in spray on/off, would the flow controller adjust flow fast enough to meet the target rate? Pottage selected an area in Saskatchewan for the test, Mackay made the prescription file and Nguyen helped Pottage to set up the Platinum for the test. Since Provincial Airways was super busy during the spray season, the first test was done mainly by Pottage whenever he found time and the weather allowed. August 21 was an exciting day for us when Pottage announced he would perform the test on that day. Steven Glas, a Provincial Airways’ pilot, flew the aircraft and followed the guidance on the Platinum lightbar to keep the aircraft on the right track. All spray on/off and flow adjustments were left for the auto spray and flow controller to manage.
The result from the first test was very encouraging. Most of the spray on/off were done precisely at the boundaries of polygons on the VRT map. There were some missing or spray over no-spray zones because the distance from spray on to off or vice versa was too short (less than 0.6 seconds of flight) as shown in Figure 2.
The AgNav Auto Spray system uses a three-way valve and a DC motor to turn spray on by opening the valve to direct liquid to nozzles; or turn spray off by closing the valve to recirculate liquid to the hopper. When there is no flow, it takes about a quarter-second to fully open or close the valve. In practice, the flow pressure and torque may cause some resistance to the valve movement. This was found in the on/off results of the first test. If the time to cross the boundary of a no-spray zone was less than 0.6 seconds, the spray continued on over the no-spray zone as shown in Figure 3. Each dot on the figure represents a position recorded every 0.2 seconds in the Platinum.
The flow controller worked better. It could maintain the applied rate to meet the target rate within 0.5 seconds most of the times when the flow was stable. Figure 4 shows the applied rate of 4.0 ± 3% GPA in blue. The last sprayed line in the south side finished only half indicates low flow rate when the hopper was empty.
From this observation, Nguyen suggested Mackay to simplify the prescription file by removing too small polygons. In practice, due to mechanical limitation the spray boom cannot be cut off then turned on within half a second. Mackay agreed to merge polygons when making the next prescription file. The team planned for the second test.
The second test aimed to try two things: First test the new prescription file made for aerial application and second, test variable application rates on the same field. Mackay also wanted to test where the actual application reaches the ground in the zones when turning spray on/off. This requires a weather station to measure spray drift from wind speed and direction during flight. Even though the AgNav Platinum could calculate wind drift and apply it for offset guidance, due to short notice it was not prepared to do this function. It was the end of the spray season, so Pottage selected an area close to his airport. Mackay made two prescription files: One for on/off test with constant rate and one for variable rate test. The variable rates changed between 0, 2, 3, 4 GPA based on zones and in-season satellite imagery. See Figures 5 and 6.
The second test was scheduled on October 04. Pottage prepared his aircraft from the day before, making sure everything was running properly. Mackay drove from his office in Calgary, Alberta to Moose Jaw, Saskatchewan to attend the test. Pottage also set up a GoPro camera to video the application. The first flight started at 11:06 a.m. local time to test the new on/off prescription. The rate was constant at 2.0 GPA. The weather was sunny but a little cold in Moose Jaw with the temperature around 48°F (9°C) and a west wind of about 10 knots. The pilot sprayed in north/south directions. The flight took about 27 minutes. The tank was refilled and the second flight started at 11:42 a.m. to test the variable rate control. The flight took about 26 minutes to complete. Both tests were successfully done. Nguyen was called to check the data.
The on/off test was excellent proving that the new prescription file worked well. Only three times spray was found over the no-spray zones. When checking closely, it was found that the crossing distance on the no-spray zones were all less than 100 feet or half-a-second of flight. All other spots had the boom spraying right on-the-money. See Figure 7.
How about the rates? In general, the variable rates worked well. The target rates were forecasted half a second ahead of time when the aircraft flew the field from one end to the other. Due to wind, spray speed varied from 143 MPH to 152 MPH. At an averaged spray speed of 150 MPH (240 km/h) and a 66-foot swath width, the flow rate was about 40 gallons per minute (GPM) or 160 liters per minute (LPM) for 2 GPA, 60 GPM (240 LPM) for 3 GPA and 80 GPM (320 LPM) for 4 GPA. The flow controller that Provincial Airways has could work in the range of 9 to 90 GPM (35 to 350 LPM). It was found that the flow controller worked the best for 2 and 3 GPA. It was able to adjust the flow rate to meet the target from boom off to on within 0.5 seconds and maintain the same rate regardless of variation in aircraft ground speed. However when the target rate was changed from 2 or 3 GPA to 4 GPA, it took up to a second to meet the target rate. There might be some room for improvement here for the AgNav flow controller when the flow rate is near the extreme value. See Figure 8.
In conclusion, variable rate treatment (VRT) technology is well-used for ground applications due to slow spray speed and controllable nozzles. However, the adoption of this technology in aerial application is slow. The results from variable prescription tests done by Provincial Airways, Bayer Digital Farming and AG-NAV proved that VRT could work well in aerial applications, as well. When making a prescription file for aerial application, to optimize VRT results, one should take into account the spray speed of the aircraft, response of auto spray system and flow adjustment ability of the flow controller to define boundaries for application on/off and variable rates.
There are no doubts about the benefits variable rate technology brings to farmers for crop yield. The challenge for the ag-pilot is to make it work with the aircraft, spray and guidance systems.