The image I included in this post represents the current approach to applying fertilizer and pesticide to a field. A tractor drives down a field and liberally applies chemicals to every plant as it goes. If you look in the foreground of this image, you can see some grasses intermixed with the soybean plants. Imagine a scenario where instead of applying pesticide to the entire field, this tractor would pass by one of these grasses and apply pesticide to the grass plant, leaving the soybean plants untouched by pesticide. In a nutshell, that’s the concept behind precision agriculture.
The benefits to this precision approach when compared to current farming methods are pretty obvious. Using less pesticide means spending less money on pesticide. It also means less pesticide getting into the soil and reduced impact on the surrounding area. Now magnify that benefit to things like applying the precise amount of water required to grow a healthy plant and the right amount of fertilizer and you can see where this starts to add up to an overall lower cost to producing plants that can feed the ever increasing world population.
Part of what makes this possible is the evolution of machine learning models and sensor technology that allows a tractor in the field to detect invasive plants. Through the application of precision agriculture, the tractor makes a decision to deploy the appropriate treatment to eliminate weeds from the field without spraying all plants.
Purdue University is doing some interesting work in this space. At HPE Discover 2018, I had the chance to talk with Pat Smoker from Purdue about precision agriculture and the impact it will have on both the way plants are grown and the way livestock is managed. You can check out the interview in the video below.
