Hack Chat Transcript, Part 3

A event log for IoT for Agriculture Hack Chat with Akiba

The bits and bytes of growing food

LutetiumLutetium 05/16/2019 at 01:070 Comments

freaklabs5:43 PM
i think data science is a huge thing for small farmers. its one of the technologies that I think is very needed

enatalio5:43 PM
@freaklabs > I like it! data mining baby!

freaklabs5:43 PM
small point of fact, data science foundations are based on farming. ANOVA was designed to analyze agriculture experiments.

sadhana5:43 PM
My area of research is mostly in CO² enrichment to counteract climate change. More recently Ive been curious with methane enrichment through pines

Yves Quemener5:44 PM
@freaklabs I can help. I wish more environmentally minded people had the data in mind.

enatalio5:44 PM
@freaklabs > are you creating anything like sci-kit for ag?

matthewkleinmann5:44 PM
I tell you, when we first moved out here, near 20 years ago now, we rarely had tick issues. Now going in the woods and picking ramps, you just pick the suckers off of you when you come back in. It is very bad.

freaklabs5:44 PM
but one of the thngs I'm obsessed with is teaching experimental techniques to farmers. I think it can benefit all farmers, indie or big ag. and its something most don't know how to do. Even a simple single factorial with one-way ANOVA that can be done in Excel could hugely boost productivity

freaklabs5:45 PM
@sadhana : not really. i'm not sure how useful machine learning is for ag at the moment. I got a suggestion that ML could help afarmer determine when to pick a cucumber

matthewkleinmann5:45 PM
Big ag has control over a lot more than small. They have it down to about as tight as you can get it, with chemicals and irrigation.

freaklabs5:45 PM
@sadhana : i had to tell the guy you don't really need to teach a farmer when he can pick a cucumber

freaklabs5:46 PM
@matthewkleinmann : yes, there is a big focus on optimization

freaklabs5:47 PM
regarding the experiments, I think the interesting thing is that each farm is different iwth different soils. So running experiments on your own farm, speficif to your growing ocnditions can help identify how to optimize for that specific farm.

matthewkleinmann5:47 PM
Like for us, current decisions are when to bring plant starts out from the house into the greenhouse and than when to put them outside.

sadhana5:47 PM
MIT has the best machine learning system Ive seen in farming

enatalio5:47 PM
@admin> Is the entire chat threads available later on?

matthewkleinmann5:48 PM
One thing we are doing this year is I build a bunch of movable containers and we are going to try and put some things back in the greenhouse near the end of the season and see how long we can eke things out of them.

freaklabs5:48 PM
ie: what conditions can you germiante quinoa seedlings in for specific temperatures. these can be run in aquariums iwth small heaters. Knowing the optimal temperature of quinoa could tell you approximately what date you can plant quinoa in your region/zone

@enatalio -Absolutely. I'll copy the whole transcript and post it in the event page -

Yves Quemener5:48 PM
@matthewkleinmann I find it intriguing that we don't have a good database for this kind of things. There are tons of local farmers almanach, but little in terms of well formatted databases

freaklabs5:49 PM
@matthewkleinmann : nice. that's ag reat idea for season extension. Actually we might have to copy you on that idea. it sounds realy interesting

enatalio5:49 PM
@Dan Maloney > Thanks Dan!

WRR5:49 PM
What do you all think about hydroponic/aeroponic greenhouses? It seems like you could control the environment and nutrients more easily, but I guess they might be expensive to make and maintain?

freaklabs5:49 PM
sorry about the divergent threads. I only have around a four chat history on my current laptop screen

freaklabs5:50 PM
ha ha ha.

Yves Quemener5:50 PM
It is 2019, your screen resolution is crazy but you can read less lines than a 1990 IRC discussion

freaklabs5:50 PM
@WRR : i think hydroponic greenhouses are really interesting. actually hydro is good for urban growing or if you need really dense growing.

Yves Quemener5:50 PM
/me goes back mumbling about kids on his lawn...

enatalio5:50 PM
@Yves Quemener > LOL!

Arsenijs5:51 PM
@freaklabs top left corner of the chat window, right above, there are two "hambugers", press the right one

matthewkleinmann5:51 PM
I will let you know how it works. The containers are small, about 1.5x the size of a 5 gallon bucket. I have a near endless supply of oak pallets and I set up a little factory in my wood shop and made them out of oak, with drainage holes in the bottom and runners you can can get a dolly under them.

Arsenijs5:51 PM

freaklabs5:51 PM
@WRR : in response to controlling nutrients, I think its one of the drawbacks of hydro. Because you only get out of it what you put in, you don't get the trace minerals or even the benefiical bacteria that you get from composted soil.

Yves Quemener5:51 PM
@Arsenijs Oh, much better, thanks!

sean5:51 PM
Yes, ty

sadhana5:52 PM
Freaklabs Id like to send you a photo later showing how crazy of a difference the germination temperature can make

freaklabs5:52 PM
Oh, the right hamburger is quite nice, but still have limited screen space. next time will use my big monitors rahter than laptop

freaklabs5:52 PM
I always though tthe name: hamburger was hilarious

matthewkleinmann5:52 PM
Re: Hydro. I have tried this on house plants with good effect, have you tried making "tea" out of rich soil and compost and trying that in your hydro?

enatalio5:53 PM
@freaklabs > actually this one is triple burger....haha

Arsenijs5:53 PM
@freaklabs blame web designers for that name

matthewkleinmann5:53 PM
It works well on houseplants if you don't wanna use miracle grow.

freaklabs5:53 PM
I'd like to experiment more with aesthetically pleasing hydro. for example hollowed out bamboo as pipes. i think it would blend the natural and tech well

freaklabs5:54 PM
but will probalby add that to the neverending pile of potenial projects

matthewkleinmann5:54 PM
Very cool and even better if you can get the bamboo locally. I use all recycled and re-purposed stuff. I am very cash poor.

Bamboo is just miracle stuff. So many uses!

Yves Quemener5:54 PM
About greenhouses, there is also a hybrid approach that works well: put sensitive seedlings in a well controlled environment and plant them in outdoor plots once they are more resistant. You don't need to grow all the plant inside a greenhouse

freaklabs5:54 PM
the unfortunate thing is that there are so mnay things that we manage day-to-day that many projects fall by teh wayside. there are literally so many interesting things to look into in agriculture

freaklabs5:55 PM
@matthewkleinmann : out here, bamboo is a pest. we can't get rid of it fast enough.

Yves Quemener5:55 PM
Oh, re-doing the aeroponics (which uses PVC) with bamboo would be so nice...

WRR5:55 PM
Cool ideas, thanks for sharing! I hadn't thought of the energy cost to produce nutrients, and bamboo does sound like a great material

sean5:55 PM
I use a radiator heater under a bunch of shelves for a 'germintation chamber'. Most things pop in 3 days. Temperature control is amazing.

Yves Quemener5:55 PM
We should talk to Noriko and Kawakami (the samurai one) about this, they would love it.

freaklabs5:55 PM
FYI, Yves is also interfacing with the company doing the weed picking robots using machine learning and ocmputer vision

matthewkleinmann5:56 PM
My biggest thing with my greenhouse is getting water to it. It is next to the pond but far from power. Thus the home brew wind mill to run a small pump that will fill a stand tank and I can water out of the stand tank.

freaklabs5:56 PM
and developing an agriculture cable-bot

sean5:56 PM

Yves Quemener5:56 PM
o/ This chat is a welcome distraction from my debugging of quaternions conversion

Yves Quemener5:56 PM
Yep, suspended robot, cable bot, wire bot, comes with many names

sean5:57 PM
ah, ty

Cable bot like they use for televising football games?

Yves Quemener5:57 PM


freaklabs5:57 PM
@matthewkleinmann : yes, I find that power is one of the big factors that don't get taken into account in a lot of ag-tech prjects. we see so many raspberry pi powered ag devices, but they would never survive in a field unless you had huge solar panels and battery

WRR5:57 PM
It looks like a giant delta printer :D

sean5:57 PM
that's what i thought

freaklabs5:57 PM
quaternions...ha ha ha. never associated that with farming

sadhana5:58 PM
Matthew the problem with hydro nutirents is that nitrogen gets different uptake in roots compared to phosphate/potassium. It needs to be consistently monitored and adjusted. ive been building an ipen source doser that detects pH, EC, and NPK ratios in real time to dose using peristaltic pumps

sadhana5:58 PM

freaklabs5:58 PM
i really like solar power but haven't explored windmill for power much. would like to investigate it more

Yves Quemener5:58 PM
I am trying to write the control in a bit unusual way in order to lure ML investors :-)

matthewkleinmann5:59 PM
I have 2 small solar installs now, Last year I built the rotor for a vertical windmill. I need to get the pole in the ground. I am going to see if I can run one or two treadmill motors as generators to run a small bilge pump.

Yves Quemener5:59 PM
About power, we also have a good hydro potential here, but it is going to be intermittent

WRR5:59 PM
@sadhana So, that seems like the sort of monitoring problem you could handle with sensors and embedded devices - is it more complicated than taking readings and adding nutrients when necessary?

freaklabs5:59 PM
In regards to running hackerfarm, I think the final result is that we're a hackerspace and a farm, but kind of more farm than hackerspace.

Jeff5:59 PM
@freaklabs @Yves Quemener what sorts of things do you like to grow? which crops are particularly challenging?

freaklabs5:59 PM
that means that almost all of our technology is in the context of agriculture or food processing

matthewkleinmann5:59 PM
Mine is both. A wide assortment of shops and ag.

freaklabs6:00 PM
@Jeff : We try to grow crops that aren't naturally found in Japan. The problem is to have things you cant find in grocery stores. at th moment, we are growing native Mexican crops like Jalapenos, Habanero peppers, tomatillos, and white native mexican corn

sean6:00 PM
@freaklabs Other than the moisture meters, does hackerfarm have any other tech that is deployed in the fields?

Yves Quemener6:01 PM
@Jeff in case of the suspended robot, right now we are evaluating where it is doable to help a lot with automation. My dream would be to be able to manually pick up pests, so that one could automate organic vegetable production (which typically require a lot of labor) but in the end it will be dictated by the capabilities and profitability.

matthewkleinmann6:01 PM
I like mushrooms, oysters and shiitakes, grapes, herbs, tomatoes, mellons, pumpkins. Critters, one pig, sheep, alpacas, bad little goats. Chickens...

It's getting on to an hour for this chat, not counting the half-hour or so of prechat discussion. I'm going to call an official end, but if Akiba is game to chat a little more, I say go for it. I'll just say a huge thanks to Akiba for making time in his schedule for this Hack Chat, which I found really fascinating.

freaklabs6:01 PM
we will be supplying mexican restaurants in tokyo with authentic mexican produce grown locally and organically.

Yves Quemener6:01 PM
But yeah, fruits and vegetables are super expensive here, due to several factors, so being able to produce some at small cost would be great.

freaklabs6:01 PM
Yeah, i'm okay to chat for a bit more.

sadhana6:02 PM
WRR, the issue is ion selective electrodes are insanely proprietary anf expensive

matthewkleinmann6:02 PM
Small fruits and most veggies do well in containers.

freaklabs6:02 PM
Thanks everyone for coming out. I know I didn't get to everyone's questions. My screen was scrolling like crazy. Please PM if you have specific aquesitons. I'd love to answer them

freaklabs6:02 PM
errr....if I can :)

Thanks to everyone for stepping outside the usual schedule for this one. And next week's Hack Chat with OSH Park is going to be off-schedule too - we'll be doing it regular time (noon Pacific) but on Thursday rather than Wednesday -