Eye in the sky
When someone tells you drones are about to revolutionize agriculture, believe them. More and more people in the farming industry believe that use of drone imagery will soon be standard practice on most major farms, helping to manage everything from potatoes to horticulture crops, grapes to field crops.
The way things are currently, growers have to scout for disease and other crop issues (and in many cases, send samples away for confirmation). But the human eye can only detect a fraction of the information that can be measured with today’s sensors. With properly-calibrated imaging tools – which can pick up a great deal of information beyond the visible spectrum – there is a whole new dimension of crop management emerging.
It’s true growers can and do use sensors to gather information now by physically walking around the fields with a sensor tool or carrying one while riding in a machine, such as the Greenseeker. But drones in the sky are more efficient. They are able to take hundreds of images across large areas in a few minutes and send the data wirelessly to a central database.
The system software analyzes data as it arrives, comparing it with norms on file, then automatically sends reports back to the grower – in some cases within 24 hours – with the time for the whole process to cycle shortening all the time.
The reports are tied into GPS positioning, so that growers can use their smartphones or tablets to go directly in the field to pinpointed areas of concern. Drought stress and a few diseases are already capable of being detected. Drone imaging is also being used to assess crop status (planting evaluation, growing stage, yield estimates), evaluate and survey for drainage, and to track weed levels.
Andy Reynolds says most remote sensing work in vineyards so far has been conducted in Australia and Europe using fixed wing aircraft.
“[It] has involved use of sensors that record red-green-blue spectral reflectance from plant canopies that provide a metric called normalized difference vegetation index (NDVI),” explains the professor at the Cool Climate Oenology and Viticulture Institute at Brock University in St. Catherines, Ont. “The utility of these data comes from spatial relationships between NDVI and yield, vine vigor, water status and possibly berry composition metrics such as sugar (Brix), titratable acidity (TA), anthocyanins, phenols, etc.”
Reynolds says there are generally direct correlations between NDVI and yield, vigor, etc. because NDVI is basically measuring greenness in a plant canopy. A low NDVI is typically associated with low vine water status, low TA, high Brix, anthocyanins, and phenols, partly due to smaller berries and the associated concentration effects.
The remote sensing work done by Reynolds and his team in Niagara in the mid-2000s illustrated these results.
“One of the challenges, which we will continue to have with our drone work which just began in 2015, is the fact that Niagara vineyards have green cover crops in row-middles that reflect the same as plant canopies,” he notes. “So, all the pixels associated with the cover crops need to be masked to provide NDVI data representative of the vine canopies exclusively.”
He names another stumbling block with assessing drone tech to be weather differences from year to year, with high rainfall seasons having the capability to completely change the zonal patterns in NDVI and other pertinent variables.
Reynolds and his colleagues saw this happen in a recent five-year study in a large Riesling vineyard in Beamsville, Ont., and a two-year study in four Pinot noir vineyards in St David’s, Ont. These two issues, he notes, are major challenges in the process of convincing the industry that the use of remote sensing – either by drones or standard aircraft – might be worth widespread investment.
Reynolds believes all this research may someday allow scientists to delineate temporarily-stable zones of a vineyard that will end up producing wine of differing quality.
“This information might be used in two ways,” he explains. “If the zones are somewhat geometric in nature, it might be possible to implement ‘precision viticulture’ whereby variable rate fertilization, liming etc. might be used to reduce the variability.
“The other way of using the data is to simply accept the fact that the vineyard is variable, and make two or more different products that reflect the variability, for example a $15 bottle of wine from the high-vigor zone and a $25 bottle product from the low-vigor zone. All this will depend upon a clear correlation between the NDVI data accessed by the drone and all the variables on the ground.”
SkySquirrel Technologies in Hammond Plains, N.S., uses drone imaging and data analysis/reporting to currently serve about 30 vineyards in Canada, Chile, France, Spain, Romania and Switzerland. The company also does a little field crop and golf course work.
About half its clients are Canadian, based in B.C. and the Maritimes, with at least one Ontario client coming on board in 2016.
SkySquirrel has a partnership with VineView of California, which started gathering aerial infrared data in vineyards using airplanes about 12 years ago.
SkySquirrel formed in 2012 and focussed in on vineyard management in 2013.
“Grapes are a high-value crop and the flight time is a good match for size of vineyards,” says Richard van der Put, company CEO.
Usable analysis is back to vineyard managers usually within 24 hours, and the flight planning training for vineyard managers takes about a day. Van der Put says the biggest challenge in developing their system was image calibration – creating software algorithms that would compare imaging data with norms and also make comparisons over time.
Cost return for drone use is best measured on a per-acre basis, in van der Put’s view.
“There’s an immediate return on investment, but how much depends on the application,” he notes. “One disease that is very important in vineyards is grapevine leafroll virus. As many readers would know, there is no treatment and you have to remove the plant. Early stage management is critical and provides a huge return on investment.”
Leafroll shows up on the images with an obvious colour change, but that could be due to other diseases, so vineyard managers usually send plant samples for confirmation testing before culling.
SkySquirrel currently has no capability to detect mildew or other diseases, but offers differential harvesting analysis (a determination of when grape harvesting should occur), and is in the process of incorporating water management services. van der Put notes that its partner company, VineView, helps vineyards in California drop water use by 25 per cent.
Drone use in hort crops, potatoes
Drones have been used in several projects in Ontario’s Holland Marsh area, according to Jody Mott, Holland Marsh Growers’ Association’s interim executive director. The projects have mostly focussed on surveying fields and buffers. Mott adds that a recent Campbell’s soup commercial involving a Holland Marsh grower was filmed using a drone.
One current Holland Marsh drone project is being spearheaded by Mary Ruth McDonald, a professor at the University of Guelph and the research trial coordinator at its Muck Crops Research Station. The two-year study, supported by federal Growing Forward 2 funding and involving Bradford Cooperative Storage Ltd., has been extended into 2016. McDonald says one objective of the research is to investigate the use of aerial photography to improve integrated pest management programs – basically to see if crop damage can be identified earlier and more efficiently with drones than through scouting (with the cost of scout labour increasing and drone costs coming down).
She and her colleagues also want to see how drone imaging can affect research itself, as it may provide more data and more objective data than that which is currently gathered.
Resson Aerospace of Fredericton, N.B., delivers data analysis for various clients, and the Financial Post newspaper reports that Resson signed a seven-figure multi-year deal with McCain Foods.
According to Nicole Rabe, a land resource specialist with the Ontario Ministry of Agriculture, Food and Rural Affairs, barriers at this stage in using drone imagery in the horticulture sector include cost, the time it takes to get usable data back, and risk.
But Rabe also asks about the risk to the grower if something has been missed. With more research projects and private partnerships looking into drones in Canada, it seems clear that answers – and a reduction of barriers to drone use in agriculture – are on the way.
SkySquirrel Technologies in Hammond Plains, N.S., uses drone imaging and data analysis/reporting to currently serve about 30 vineyards in Canada, Chile, France, Spain, Romania and Switzerland.