Since I accepted the role of leading Research and Development for the Crop Science division of Bayer back in June, I have been visiting our R&D facilities all over the world. I really enjoy getting to know our incredibly diverse teams and seeing firsthand some of the exciting agricultural innovations each site is working on.
The only downside: months of travel means many nights spent in hotels.
Whenever I travel, whether on business or vacation, I always try to find the hotel that best meets my personal needs. Centralized location and reasonable cost are always considerations, of course. But I also like to have access to a gym for my early morning workouts and decent dining options within walking distance. I could search hotel websites individually to do my research, but it’s much more efficient to use a travel website that aggregates information and guest reviews from multiple hotels in one convenient location. By accessing data from many other travelers about their experiences, I can make the most informed decision for myself.
This same idea of crowdsourcing data can help farmers as well.
Farmers have been collecting data about their operations for hundreds of years – first in paper ledgers, then on computer spreadsheets. They have recorded details like what types of seed varieties they planted and when; how much and how often they applied each type of pesticide; how much rainfall their fields received; and ultimately, how much yield was harvested from each acre. The problem is farmers have never had a good way to analyze all this data. And data itself doesn’t have any value unless it provides insights that help you make better decisions.
During my recent visit to the San Francisco headquarters of The Climate Corporation, a Bayer subsidiary, I learned more about how today’s modern ag technology is finally able to capture some of that value.
Now, all a farmer’s data can flow seamlessly from sensors, satellites and field equipment into digital tools like Climate FieldViewTM, where it is compiled with publicly sourced information on weather, soil types, topography, etc. FieldView can combine multiple data sets and filter them through various algorithms that help growers decide what products to use, in what amount, in what location, and at what time during the season.
For example, a farmer could find out which hybrid crops have performed best in fields with sandy soil, and within that set, what other growing conditions were present in the most successful crops. FieldView also analyzes data to help farmers use nitrogen fertilizer more efficiently, thus reducing potential impacts from associated greenhouse gas emissions and water runoff. And all that data is visualized in digital maps and charts that are easy to understand, on a smartphone or tablet that the farmer can access anytime, from anywhere.
But learning to effectively harvest data, across any industry, is not without its challenges.
One challenge is simply helping growers understand that there is strength in numbers. Rich data sets that come from several applicable sources are more useful when used together. Smaller data sets are statistically not as accurate, meaning they are not as helpful when driving decisions. How do we get farmers used to the idea that one piece of data isn’t nearly as powerful as tens of thousands?
I have heard some colleagues illustrate this concept using digital cameras. When the first digital camera was released, it captured less than one megapixel per image. Today, most smartphone cameras capture around 12 megapixels. Think about the difference in photo quality between the first digital camera you ever owned and what you use today. The older images may be too grainy to print at all, while you could probably print poster-size copies of the images you’ve taken with your current phone. Each pixel contains data, and more data gives you a sharper picture. Likewise, with each acre added to the FieldView database, Climate’s predictive algorithms get stronger, which means better insights for each grower.
If the industry can navigate these challenges successfully, I have no doubt that harvesting data will indeed lead to better harvests.
Advancing data science in agriculture will help farmers gain deeper understanding from their fields – so they can optimize each field’s specific genetics and unique yield potential. And that, in turn, will benefit us all – helping farmers produce enough to feed the world’s growing population more efficiently and more sustainably than ever thought possible.