Build a Personal Food Computer | Hacker News Reader

Build a Personal Food Computer

openag.media.mit.edu

81 pointsTrisell43 comments

Farmers would call this a "greenhouse". Here's one commercial greenhouse control system.[1] There are lots of greenhouse control systems. The fancier ones control water, fertilizer, temperature, humidity, oxygen content, CO2 content, lighting, etc. Some vegetables, such as lettuce, can be overclocked, with a short day/night cycle.

This stuff isn't new, although the controllers are getting smaller.

[1] http://www.climatecontrol.com/

Are any of them built on open source software and open hardware ? I think that's part of the big idea here.

One could, eschew software, and handle the requirements and growing manually. It's actually pretty rewarding when you have the time and space.

This strikes me as a non sequitur. If I wanted to eschew software, I wouldn't be reading about building a "personal food computer."

https://www.reddit.com/r/SpaceBuckets/ A community of 20,000 DIY indoor gardeners.

Subset from the SpaceBucket manifesto:

Humans of the Earth! Hear the Space Buckets calling. It is time for a DIY indoor gardening revolution. We stand at the vanguard:

    I. Bucket culture is open. We are a community of learners, a movement of tinkerers. Our goal is to spread information about how to grow plants inside small containers with artificial lighting. Anyone can join us.

    II. Bucket knowledge is free. We believe in the free flow of ideas and the unparalleled power of the Internet. We are part of the web, and we contribute to it. Profit is not our motivation, innovation is.

Also found these:

[1] http://hmeter.sourceforge.net/

[2] https://www.3dponics.com/ ( a 10,000 person community )

[3] http://yieldbuddy.com/ An open source automation system for hydroponics

[4] http://www.powerhousehydroponics.com/crash-course-open-sourc...

[5] http://scienceinhydroponics.com/2013/02/the-first-free-hydro...

Are any of them built on open source software and open hardware ?

Yes: http://opensourceecology.org/wiki/Aquaponic_Greenhouse_Works...

Farm Hack is another community of farmers collaborating to create Open Source tools for Ag. http://farmhack.org

> Some vegetables, such as lettuce, can be overclocked, with a short day/night cycle.

Do you know what this is called?

Traditionally (19th century British literature) any such technique was called "forcing" if it simulated an environment more favorable to growth. Techniques include light cycle variation (covering the plant with an opaque cover at specific time was the standard approach), thermal environment manipulation (using horse dung to generate heat from below, or glass cover to trap heat above), CO2 manipulation by breathing in the space the plants are, or using bacteria that produce CO2, varying the timing, location and temperature of water, etc etc etc

Here's a photo of people harvesting forced rhubarb. Someone said this was in Cheltenham (Gloucestershire, which is south west England) in 1937.

http://imgur.com/B8iqwAh

I don't know anything about this photo. There's a lot of information that's been lost - I don't think anyone knows who the people are, or where this was taken, or when it was taken. Maybe the information is in some archive somewhere and it just hasn't been digitised yet. It does make me wonder what the future people will call this time - the time where we have photographs and film and audio, but it hasn't been digitised and is hard to search.

So we don't know why this women is wearing nice shoes while picking rhubarb - did she just not care? is this a special occasion photo op?

Sure - a lot of things aren't digitalized, but that doesn't mean it's lost - yet. A lot of it still exist in physical and analogue form - in archives, collections and private as well as public libraries across the globe.

There are however some quantity of people who equal "not being digitalized" as "being lost" (I'm not saying you're one of them). That would however not be true, at least not yet.

That said, there are people who have an over-confidence in technology or digitalization, which think we should forgo our analogue copies of any or everything. Some people use the argument that "Everything of value, has been digitalized". (Not everyone is a curator... tho)

I'd be interested to see the difference in nutritional content (macro- and micronutrients) in over clocked lettuce vs regular, and other similarly hacked foods.

If you were to find significant variation (e.g. lower nutrient quantity/quality when using overclocked growing methods), it would be interesting to see if you can pull the needle in the opposite direction, e.g. "underclocking" to try to create more nutrient-dense variants.

Haven't you learned? We just give things new names and people are amazed.

Statistics and vacation rentals aren't new. But when you call it "data science" and "sharing economy", it sounds cool.

It's a "3D printer for food"!

https://www.cayneshousewares.com/skin1/images/Products/34533...

From another page about the project:

The Food Computer is a controlled-environment agriculture technology platform that uses robotic systems to control and monitor climate, energy, and plant growth inside of a specialized growing chamber. Climate variables such as carbon dioxide, air temperature, humidity, dissolved oxygen, potential hydrogen, electrical conductivity, and root-zone temperature are among the many conditions that can be controlled and monitored within the growing chamber. Operational energy, water, and mineral consumption are monitored (and adjusted) through electrical meters, flow sensors, and controllable chemical dosers throughout the growth period.

Each specific set of conditions can be thought of as a climate recipe, and each recipe produces unique results in the phenotypes of the plants. Plants grown under different conditions may vary in color, size, texture growth rate, yield, flavor, and nutrient density. Food Computers can even program biotic and abiotic stresses, such as an induced drought, to create desired plant-based expressions.

The name 'food computer' is probably better used for the computer / control portion of their system. Reading their description, 'food factory' or 'automated greenhouse' seem like better names.

I am wondering how long we will continute to use plants and animals for food production. From an energy perspective, I see food as digestible bonds and I am sure there are much more efficient means of creating chemicals with these bonds that are safe. Fully synthetic foods will enable the next chocolate (in terms of experience), won't require nearly the amount of energy/land/labor to produce, won't harm plants or animals, and will be able to be ultra customized since they won't be derived from biological sources.

I think in the long term many people will choose just to grow their own food, in a very conventional way, on a small scale, coached by an AI. There will probably be a bit of a fork: some foods going the way of just being 3d printed from commodity traded ingredients and sold basically at cost of ingredients, and other food that is completely made by hand by a person, in a totally unique batch. And to be honest, everything in between.

Which customers will choose will depend on whether they want a food experience, or they just want some material item? Or they want a little of each.

Synthetic materials don't come from nothing. Following the supply chain backwards they're all[1] mined from somewhere, either as fossil fuels or minerals like potash and limestone. All the NPK fertilizers (which this system uses) come from mined materials.

Far from "not harming plants or animals," this material extraction comes with huge (but externalized) ecological, social justice, and wealth concentration problems. And of course it's completely unsustainable.

Now compare that to how healthy soil functions: microorganisms leach minerals directly from grains of sand/loam/clay, essentially manufacturing in-situ minerals into bioavailable form right at the root zone. Nitrogen fixation is similar, performed by leguminous plants using otherwise wasted sunlight and ISRU. https://www.youtube.com/watch?v=x2H60ritjag

[1] Since I know someone will bring it up, some do come from plants like soybeans. But mechanized soybean production is primarily either soil mining or (in areas with badly degraded soil) entirely supported by fossil fertilizers.

Somewhere or other I heard about a guy named Masunobu Fukuoka. He's written a book called the "One Straw Revolution". He was a plant biologist who decided to research very low input farming (not "organic", actually virtually no inputs other than water). His basic premise was that if you tried to optimise how nature handled the crops rather than trying to do it yourself as the farmer, you would get good yields with very little work/cost. He famously suggested that each person should be given a quarter acre of land and that they should spend an hour or two a day using his "natural farming techniques" to feed themselves.

There is some controversy over whether he really got the kinds of yields he reported with the kind of work he said he did. Part of the problem was that he got really famous and people would wander over and help him out for a while so that they could learn his techniques. However, I have a friend who studied with him and I can report that he literally does manage to feed himself and his wife all of their fruit and vegetables and only spend some of his spare time (he is a high school teacher by trade). The biggest piece of labour is cutting the rice, which takes several days (and which he does all by hand).

When I visited his fields (in Japan), it had been a bad year that had been compounded by a typhoon just before harvest. My friend's field was in great shape while you could see that the neighbours had lost at least 20% of their crop. It was also really interesting to see the difference in wildlife diversity between the fields. These fields were very small (about the size of a large house lot in the US) and my friend's lot was full of frogs, snakes and spiders, while the neighbour's had nothing in them but rice. My friend said that his neighbour usually produces more rice than he does, but that his field is always much better in a bad year.

I actually did a 5 year experiment trying to grow food in containers with no inputs other than water. You need fairly big containers (depth is the most important thing), but I was surprised how well it worked. At the end of the 5 years I was getting pretty good yields with virtually no work and no inputs. In the end I had to give up because while the containers grew a lot of food, they also grew a lot of weeds. The neighbours didn't understand and one day I got home to find the gas man spraying herbicide on everything. Everybody thought it was an eyesore and he was delegated to "fix" the problem. :-P

In any case, my experience makes me think that there is room for some research in farming where we don't try to micro-manage everything.

Where do you live that spraying herbicide on someone else's garden isn't considered tresspassing and destroying private property?

Maybe it is, but on the other hand he knew his "garden" (containers with weeds in them) were a nuisance to the neighbors. Litigation is often possible while at the same time not really the coolest thing to do.

I have no idea why you were down voted because that was exactly my reasoning :-) I live in rural Japan. I admit that while I was really angry about someone just coming in and pouring herbicide on my plants, I was more upset that someone didn't just come and talk to me about it. But that's Japanese culture and in retrospect I should have been a bit more outgoing in trying to explain to people what I was doing. In the end I screamed at the gas man from half way across the neighbourhood and he was so embarrassed that he ended up having to transfer to a different part of the city. It probably sounds strange, but I still feel really bad about that. He ruined my 5 year long experiment and killed my plants, but at the time I was yelling at him, it was already done. Hurting him in order to satisfy my anger doesn't actually improve the situation.

Edit: I should point out in my own defence that I didn't actually realize ahead of time that my planters were a nuisance. I weeded out all noxious weeds and my containers were fairly well isolated with about 10 feet of concrete in every direction. Also, there was a parking lot on one side of my lot and an orchard with about a foot of mulch on the ground on the other. The spread of weeds should not have concerned anyone. I honestly think they thought I had just abandoned the planters. It's too bad, but a lesson for me in the importance of optics when you do hair brained experiments at home ;-)

> Following the supply chain backwards they're all[1] mined from somewhere

What about using (all kinds of, including human) waste as source material? We're able to genetically engineer bacteria and fungi, these could be used to convert anything biologically degradable (which includes plastics) into raw compounds which can then be re-synthesized into edible "food"... just won't look the same as "the real thing".

Not sure I buy these arguments. Plants are mining the ground for minerals it just takes tons of water, fertilizer, plant optimization, pesticides, transportation, harvesting etc.

The question is really can a solar panel + responsible chemical process provide a better product with lower energy and lower environmental impact. This needs to be measured in CO2 output per calorie, m^2/calorie, pesticide/poison introduction etc etc.

Its amazing how quickly you would dismiss this as it seems like an incredibly complex question.

Associated TED talk that turned me onto this project.

https://www.ted.com/talks/caleb_harper_this_computer_will_gr...

I had a really hard time with the "Food Computer" naming before I watched Caleb Harper's TED Talk. After watching it, I get it.

Unfortunate that the HN link is to the 'build page' and not to the front page, which has that TED Talk. Thanks for linking to it.

Dan from OpenAG here, link to github for those interested.

https://github.com/OpenAgInitiative

Dan is there anyway that we can contribute to the project? I know that you guys are working to actively move forward, yet there still might be people who are interested in contributing?

Personally I'm going to go out and build a modified version of this to see if I can get it to grow a decent variety of plants for my family through the winter in my garage.

We are absolutely interested in contribution! We are still setting up the infrastructure but our hope is that people building our V1 system or systems inspired by the V1 will share information and advice. Additionally, if there is something about our HW or SW that you don't like, create a mod, document it and share it with the community.

If you were to build your own version, it would be great if you documented the process including material dimensions, fasting techniques, different sensors or electronics, as well as successes or failures in trying to grow something. We have software that allows you to control sensors and actuators with a raspberry pi and Arduino mega. Once some or all the electronics are hooked up, the box you put it in is really up to you.

Future versions will try to standardize how we attach things, but for right now my impression is that you want to get up and running ASAP.

So far, the people building our V1 chat with each other in a Slack group. So we can add you to that one if/when you are ready to start building something.

Something I would like to see are systems for identifying plants and handling them automatically. Eventually combine this with versatile robots that pluck harmful plants, harvest exactly when ready and maybe even attack harmful insects and you might be able to replace chemicals and back busting manual labor. Farms can become food prairies where tons of beneficial plants grow mixed together, mutually strengthened against weather and disease with robots tending and harvesting throughout the year.

After working in the hydroponic and aquaponic industry for 3 years, I can say that automation for removal of bad plants is the last key needed to create fully autonomous farms. The margins in farming aren't that high and the labor for harvesting(already solved by robotics) and separating the good from the bad plants are the most time consuming and costly tasks. Having that automated would allow you to have just one or two technicians on hand in a farm rather than a handful of technicians, a lot of laborers, and a manager.

Edit: by technicians I mean farm technicians, the people that have a background in plant ecology or are botanists

Now combine this with aquaponics...

https://en.wikipedia.org/wiki/Aquaponics

This looks interesting and might be good for the future human. However, I somehow feel it is a bit sad to make agriculture so industrialized.

Yes, I feel sad about it too, though most seem to respond to this that it's a good and exciting thing.. Reading One Straw Revolution is the best way I can explain the feeling.

Just what I wanted to see rural America displaced by -- dusty shipping containers stretching to the horizon, each with PV on top and a Monsanto logo on the side...

Don't worry though. This fills a niche similar to current-gen 3D printers: extremely useful for rapid R&D, but despite the hype simply too expensive and energy-intensive to undercut conventional production on a mass scale.

What do you mean? Compared to how it is now?

I somehow feel it is a bit sad to make hunting and gathering so industrialized with all of this "agriculture" you're talking about.

This is how carrots are harvested: https://imgur.com/gallery/AP4x35k

You know those cute pictures of people picking fruit on your orange juice bottle? Nope, this is how we get oranges: https://www.youtube.com/watch?v=Av17eM1Ruyo

Of course a lot of farming is simply done by mass labour, which works if a little inefficient. On the high tech side we have GPS-controlled tractors and farmers can order up hyperpsectral images of their fields to check crop performance. Modern agriculture is extremely heavily industrialised to satisfy the demands of the billions of hungry people around the world.

This is really interesting. Thank you.

I'm not sure how literally the OP meant the word "industrialized," but to me it seems like the goal of the PFC is actually less agricultural industrialization. If it's sad that a computer is helping to grow your food, ok. But I think the point is that a local farmer could improve their crops, or you could even grow food in your own house.

I think the concept is awesome and I'm totally up for building one, though having looked at their Github it's pretty rough and ready right now.

On the one hand I think there is something cool about old school gardening and actually growing your own stuff. On the other, the geek in me wants to hack the hell out of it. One of the major problems I have every time I want to grow salad is that it gets destroyed by snails and indoor gardening solves the pest problem quite nicely. Being able to download an ideal growth profile for a particular seed would be great.

Reply on news.ycombinator.com