While it's cool and all to want to work on AI+Robots, because obviously that's where this thought is coming from because who wants to actually work in the fields am I right?\s
I think it's more important to think about market fit and margins of farming, in what world is introducing a complex, expensive machine to an already thin margin industry a practical solution?
The reason it's inefficient and wasteful isn't because "they're some hicks who need the righteous engineer to come and save them with their brilliance" its because its fast and cheap
Modern industrial-scale farming equipment is already highly engineered. Adding features that allowed for more precise care of individual plants would be an extension of existing technology, not something radically new. Take a look at modern combine harvesters, for example:
This is also a sector that would benefit from electrification, incidentally.
Of course, the fossil fuel and petrochemical sector won't like being cut out of the agriculture business to any extent, so there's that kind of resistance to such changes.
Yes, but fast and cheap is temporary (and the end result is a lot of pain).
This is a problem that requires structural investment from government (the populace) but once done, the return on investment seems very high.
How do you think the Netherlands are the second leading agricultural exporter in the world right now? Short answer - very efficient greenhouses.
To get an idea of just how efficient - they use ~4 to 9 liters of water per pound of tomatoes produced. The world average is 60 liters per pound of tomatoes. Not the high end - the AVERAGE.
In places where you have enough water, that doesn't sound like a problem.
It's not like the water is lost forever. (In those places where you have enough water, and don't take it out of an ancient non-replenishing aquifer or so.)
Sure, but water is hardly the only input they're optimizing for.
They're reducing the need for pesticides and fungicides by controlling the ambient environment around the plants.
Reducing the need for fertilizers by optimizing growing conditions and limiting runoff and waste.
Basically - the attitude there has been: "Build the right environment for the plant" followed by a focus on efficient (and therefor cost effective) inputs.
We aren't building the right environment, we're just dumping inputs (pesticides, fungicides, fertilizers, water, etc) on the growing area and calling it a day.
Which is cheap on the up front capital costs, but much more expensive over the long term, as you end up needing consistently more inputs over a long period of time.
---
Basically - I'm arguing that the up front capital costs likely are worth the returns to you get in efficiency, but farmers in the US are not incentivized to make those investments. Or perhaps more realistically - can't afford those investments on their current operating margins. Which is why we likely will want government programs focused on this.
> They're reducing the need for pesticides and fungicides by controlling the ambient environment around the plants.
Not just the ambient environment. When you go soilless through hydroponics, you eliminate a lot of soil borne disease issues. Even soil control through pots instead of direct ground contact prevents a lot of problems wiggling around.
AI robots that do companion planting with robots. There's a free idea for you. Harvesting would not work with combine harvesters. You'd have to harvest corn planted with beans like you were picking strawberries, but with robots, maybe it's doable.
Maybe, but that's a lot of degrees of freedom of movement and machine vision to use "3 sisters" farming techniques, interplanting works because of the system of moisture, structure, and nitrogen fixation that each unit provides. Not saying it's not a good system/cool project idea.
