Posted on October 14, 2014 @ 04:22:00 AM by Paul Meagher
In my research on the topic of weeds, I came across some interesting ideas about how the process of weeding might be automated. It is difficult to imagine how it might be automated, but there are some interesting attempts to see into the future. I want to share two of these with you today.
The first vision of automated weeding (see Future Directions for Automated Weed Management in Precision Agriculture by Stephen Young, George Meyer, and Wayne Woldt) sees a rover-type vehicle with a body that straddles the ground and contains different weeding devices that cultivates the soil between the rover tracks. In this depiction it has a flame weeder, a herbicide applicator, and two different mechanical devices for physically removing weeds. While the rover is going through the field crops, it is also communicating with an overhead drone perhaps to remember where and when it has completed an area and allocating it efficiently to those areas of the field that need the most immediate attention. In this vision of agriculture, the selling points are automation, precision, efficiency, and data gathering and analysis.
Not only is the vision compelling, with an air of inevitability about it, but the market for such technology could be large. Current visions are for these rovers to manage field crops occupying huge tracts of land. I wonder, however, if it will be the home garden where the technology is launched first? Why? A few reasons, but one would be battery life. It could be that we can offer batteries cheaply enough that the rovers could take care of the weeding chores at the scale of a home garden. You might be sitting in a lounge chair, sipping your favorite drink while observing your robot gardener working through your specially designed garden to help grow some of your favorite vegetables and flowers. Or maybe you are a serious gardener, a market gardener, or a disabled gardener and the agricultural robot is your best buddy helping you with a garden beyond your personal ability to manage.
These speculations seem too futuristic and perhaps not worth thinking about. We do, however, have a working agricultural rover prototype, Prospero the Robot Farmer, that successfully planted a corn field. In this video by David Dorhout, the developer of Prospero, he suggests a paradigm shift might be coming to farming.
David's idea for universal robot farmers communicating with each other is very interesting from a research and development perspective, but I think the immediate industrial market is for more specialized robots - robots that can pick grapes, apply herbicides, harvest lettuce, etc...
I do not know where the future of agriculture is headed, but automated agriculture is likely to be one of the significant directions. We already have plenty of automation coming into agriculture in the form of automatic milkers, automatic feed delivery systems, self-driving tractors, and so on, but these on-the-ground agricultural robots might be a new market opportunity given the rapid advancements in the relevant technologies (e.g., robotics, battery storage, pattern recognition, ai, precision agriculture).