Robot Agriculture
One route on my way home takes me down a road that crosses fields of strawberries. What I’ve learned driving down this road is that good strawberries are hard work. It takes a lot of backbreaking effort to comb through the rows, pruning weeds and pick fruit. There’s also the planting effort, preparing the fields, and all of the other work involved.
At the turn of the 1900s, Fritz Haber helped develop the Haber-Bosch process for production of ammonia. This allowed for fertilizer manufacturing at scale and a huge increase in the yield of fields. Before that, we were limited to natural deposits of guano that needed to be shipped across the world.
We’re likely reaching two tipping points that will help scale agriculture in the same way artificial fertilizer did a century ago. The first is genetic engineering that will help reduce waste and increase yield through pest and drought resistant strains of plants. The second is robots. Robots will remove the need for back breaking work.
The promise of robotic agriculture has been around for awhile but one of the limiting factors that was solved was cheap and reliable computer vision. A robot can identify strawberries and computing is cheap enough for a robot to run path planning.
Just like the laundry folding robots, agriculture robots don’t necessarily have to start by working fast. They can take their time and be reliable. The result will be the ability to scale agriculture even further and potentially allow for customized plants on demand that wouldn’t be possible in today’s systems.
