An exclusive interview with an expert in Western Canada, where many precision ag companies have been launched.
Western Canada – home to millions of acres of cropland – and the birthplace of many precision agriculture companies.
For a look at how precision ag technologies used there are evolving, Future Farming secured an exclusive interview with Dr. Steve Shirtliffe, Professor in the College of Agriculture and Bioresources at the University of Saskatchewan, who specializes in crop phenotyping and imaging through drones and is actively pushing the boundaries of precision agriculture in his lab. All the grad students in his lab have some coding ability but there is also a full-time programmer.
Shirtliffe has also co-created a brand-new Precision Agriculture Certificate at the University, of which he is the primary instructor. The first courses for the Certificate started in September 2022.
Why in particular has Western Canada lent itself to the development of precision ag tech?
“If you look at the landscape, there is a lot of geographic variability here. We often have too much water in the spring but often too little water in the spring and droughts in the summer and fall. Overall, we’re a semiarid area but also very fertile. So, it’s an area of the world where there is a lot of opportunity to see if you can make technology to deal with these issues and really make a big difference to yield, to save farmers money and improve efficiencies.”
Precision ag is still in its early phases. Technology is still developing by leaps and bounds
Where is precision ag at these days, in your view?
“We are still in the early phases. Technology is still developing by leaps and bounds. We’re seeing some of the approaches that companies initially used for satellite classification into management zones now evolving very rapidly. This is built on the work that resulted in today’s very advanced products that can accurately forecaste yields on a spatial basis.
What really propelled that is the whole revolution in big data, driven by cloud computing platforms. In many cases, they have access to multiple decades of satellite imagery that is available for public or companies to use. For example the Landsat family of satellites was first launched in 1972 and by the early 80s they started imaging on 30 m by 30 m pixels and imaging every time the satellite went over. That archive is entirely available.
The Europeans have also really led recently in satellite deployment. In the past seven or eight years, they have released the Sentinel family and with Sentinel 1 and 2, we now have much more frequent higher-resolution images. Sentinel 2 is 10 m pixels, and those images are freely available as soon as they are captured. This is matched with local weather data on precipitation, temperature, wind, humidity, not just on that day but all the year long and historically as well. This is a huge wealth of data now available to farmers.
Then it’s matched of course with farmer data and with the soil side of it where people are calculating, looking at the topography, where the water will go and how long it will likely stay depending on soil types. We can measure the electrical connectivity of the soil which can give us ideas about how saline it is, which is a serious issue anywhere, and how much moisture is in the soil and or its potential moisture-holding capacity. We very much need more automated soil sensing technology, however.”
We still need simplicity in any technology. The industry understands this and is constantly designing systems that have more user friendliness
What are the overarching factors at play in what’s to come?
“What’s already in development is more accurate and more sophistication. But I think we still need simplicity in any technology. The industry understands this and is constantly designing systems that have more user friendliness.
Also, the science is starting to catch up. Precision ag tech has been based on models and there are now some better models but more validation through on-farm testing needs to be a larger part of precision agriculture. If you can design methods that allow experiments to be automatically implemented in fields and automatically use that feedback for improvements, that’s the way you will have rapid progress. You have a system that teaches itself. This isn’t going to happen tomorrow. It’s small steps that are going to come together with a lot of hard work.”
Please tell us a bit about the Precision Agriculture Certificate.
“The Associate Dean (Academic) of my Department, Professor Fran Wally, and I started talking about it years ago. At the precision ag conference that’s held here, pre-Covid, we did a survey and then we invited members of the industry to give more feedback in essentially a working group. Overall, we found industry wants students to have basic agronomic knowledge and a multidisciplinary skill set. We want the students to work well in teams and communicate well (soft skills) and also have the technical skills.”
What is your own main goal in terms of pushing precision ag forward?
“In the next few years, I want to be able to create an accurate map for every field of Western Canada that would describe the inherent spatial variability (elevation, soil type) and correlate that with average levels of crop productivity.
That would tell us where it makes sense to apply spatial variability to crop productivity – where the elevation differences are such that they warrant variable rate fertilizer application practices in different areas of the field in order to reach the true yield potential, the best disease control and so on. There will be areas where it’s clearly advisable, soil moisture present, to apply more fertilizer to reach higher yields, some areas where it’s not and then the in-between ‘maybe’ areas. We’re going to probably have to dig deeper to figure out in those ‘maybe’ areas what fertilizer rate is best.”
What are your top three predictions for precision ag by 2025 in terms of what technologies will be available, what improvements farmers will be able to see, what will be automated?
“I think we’ll see much more accurate predictability of how yields vary spatially within fields and we are also going to see this accuracy in predictability being honed during each growing season as it progresses. They are making good progress with this in Australia.
I think we’ll also see more advancement with selective spraying of herbicides and technologies to measure soil nutrient levels.”