I was reading about direct air capture technologies today. The idea is really cool (passively removing carbon from the air), but they are also hard to make. You can’t just go in your backyard and build one.
For one thing, you need to do something with the carbon, usually captured as a gas, in order for the machine to be worthwhile. Typically, the companies that build them try to sell the carbon as a gas for products (think fizzy drinks) or turn it into fuel. Not exactly something a tinkerer like me wants to get involved with.
But as I thought about it, I realized that the concept is not interesting because it’s explicitly removing carbon from the air. It’s interesting because it’s a passive way to reduce humanity’s carbon footprint. The article helped make that point when they said that a flight produces emissions that could be offset by direct air capture.
Even if they just bury the gas captured from the air, they can offset emissions created elsewhere. So it’s a passive way of offsetting emissions. And that got me thinking; ok, so I can’t really inject carbon into rocks beneath the earth, but if I could replace emissions created somewhere else, that is really the same thing. A passive technology, that I can build myself could replace a more energy intensive process elsewhere.
Ok, now we are cooking. But where is there an energy-intensive process that I could replace with a passive technology? Land-use!!!
I had read another article, months ago, about how fuel efficiency standards in the US mandated the mixture of more bio-fuels into gasoline as a way to offset emissions from vehicles. The idea was that if more of what you burn comes from plants (not fossil fuels), then the act of driving is less carbon intensive.
The oversight in the plan was that growing things takes land, a lot of land. So what ended up happening is that developing countries clear-cut their rainforests to create enough land to grow palms and produce palm oil. We got our bio-fuels for gas, but the net effect was to deforest so much land that the net effect was probably negative.
The key issue here is that there is only so much land in the world, and if land is used for solar panels or bio-fuels, that land is no longer available to grow food. And we need food.
So let’s circle back to the original goal: create a passive technology that replaces a more energy intensive process elsewhere. We know that we need land to grow food, but if we were to grow food with less land, we could use that land for other things, like producing energy.
So now we are getting somewhere. Can we create a passive technology that produces food on less land than is typically used today? If we could do that, then the remaining land could be covered in solar panels or put to some other energy generating use. Here is the problem: I don’t have access to farmland. I live in a suburb, the only farmer I know lives in Illinois, and I don’t think the local authorities would look kindly on my experimenting with the sports fields in my area.
Solution: instead of shrinking agriculture’s footprint in an existing agricultural setting, what if I opened up new areas to agriculture? In fact, I do have access to some land that is not being used for agriculture, but could be: my backyard.
Part Two: The concept
Greenhouses are pretty cool; no matter what your natural growing season is, a greenhouse can be used to extend it. And because they tend to be enclosed, a sufficiently tech-ified greenhouse can be fairly self regulating in terms of temperature, humidity, etc. The other intriguing thing about the self-enclosed part is that you could theoretically introduce enough sensors and robotics that the greenhouse could be self-operating.
Imagine that: a greenhouse that notifies you when your fruits or vegetables are ready. If it was built in a way that it could power itself, collect its own water, grow its own cover crops (for fertilizer), pick its own weeds, and harvest its own vegetables, you are pretty much at the finish line for a passive technology.
In terms of unit economics, imagine that you sold or leased enough of these greenhouses to replace the vegetables grown on a small farm. All of a sudden, you could buy that farm and convert it to some sort of energy production facility. The vegetables are still being grown, but on new land, which makes the old land available for something else.
We just solved the land-use challenge of climate change! (And that doesn’t even account for the labor replaced or the transportation costs removed because the food is grown feet from the proverbial table.)
Part Three: The robot
Technically, what would be needed to make a passive greenhouse. (This is going to be a stream-of-consciousness list for now.)
Water – You need a way to collect water for irrigation, ideally from rain. That means the green house needs to have some sort of catchment, as well as a container to store the water and a pump to get it where it needs to do. From a sensor perspective, it would need to have a way to alert the owner when more water is needed and a way for them to add it.
Dimensions – I feel like a longer, taller, thin shape would be ideal. Think of it like a wall that can be put somewhere in any old backyard. Maybe 4 feet deep, 8 feet tall, and 8 feet wide. That’s not much bigger than a backyard shed. And depending on the design of the robotics, you probably don’t need the full 4 feet of depth. At the end of the day, you should be able to end up with many shapes that could be customized based on what the owner wants to grow.
Structure – The structure of the greenhouse probably needs to be wood. Wood is good because it can be sourced locally too. A greenhouse needs some sort of shell that lets in sunlight; glass ideally, but that could be expensive. It would be interesting to experiment with plastic sheeting. I’m not 100% sure about how much light needs to get in, but the structure probably only needs to have translucence in some areas and not others. Lastly, you need some way to anchor it in the ground; ideally not with concrete, but something that will prevent the greenhouse from tipping over.
Sensors – Basics include heat and humidity. You need some cameras to examine soil health and look for any issues around disease. The outside of the structure probably needs some sort of weather station to detect external temperature, rain, wind, etc. A bunch of soil moisture sensors would be needed to pinpoint irrigation. A pH sensor feature would be really cool if it could help apply the right fertilizers.
Power – The question here is really how big of a solar panel would this require. A basic 4 ft x 2 ft model or could you get away with something smaller. Would having a basic windmill make sense (probably geography dependent)? How big of a battery do you need to include?
Farming – I don’t know much about farming, but I know that for this to be truly viable, you need to be able to plant seeds, pull weeds, harvest vegetables, remove old plants, replant cover crops, harvest cover crops and use them as fertilizer. And I’m sure that I’m missing a lot. And again, for this to be viable, I suspect all that would need to be done by robotics. Fun.
Structure Automation – There is going to be a trade-off here around how much space this thing needs and how structurally sound it’s going to be. If the greenhouse is going to take up little space, it needs to be taller rather than wide. But tall things are more vulnerable to falling over, especially in the wind. So an ideal design would be one that is tall and doesn’t take up much square footage, but also has the ability to protect itself from wind. That means enabling wind to go through the greenhouse if it gets too windy. You could imagine that being based on the wind sensor detecting sufficient wind and then the automation opening flaps to allow air flow or even changing the shape of the structure to account for wind direction. Last thing you want is a sweet design that can’t stand up to Mother Nature.
Robotics – There are two aspects to this: an oversight function and a manual labor function. The oversight function is probably more about checking soil conditions and applying appropriate water. I see this as a set of sensors with a water tube attached that can “visit” each plant, pick weeds as necessary and apply water where needed. This one may need to have the ability to pick vegetables and pick them on a conveyor belt. The manual labor part of the process is probably more about removing old growth. Imagine removing the tomato vines or harvesting cover crops. This would be a more torque-heavy device that probably has a saw involved too.
Construction – In terms of unit costs, there will probably need to be a robot assistant to help construct the thing. The basic tasks of building the frame and putting up the plastic could in theory be automated. Some of the sensor work and interior robotics probably needs to be pre-assembled and possibly installed by hand.
Part Four: Let’s do this
So yeah, that would be cool. Making it fully automated is the real challenge here. If I can make it so that a user has to do nothing other than collect their vegetables when told, I suspect this thing could be viable. If it’s any more work than that, I’m not sure how much of a market there is.
But then again, for a tinkerer, that is why it’s a fun challenge. Want to pre-order one?