Ideas for organic computer? (based on pH of fruit juices)

[Jason Bobe]

Kelly Andres is working on a fruit computer laboratory

http://wiki.medialab-prado.es/index.php/Fruit_Computer_Laboratory

She's looking for feedback:

<snip>
i was wondering if you have any ideas for
a project i am collaborating on at the garage science workshop at media
lab prado. Here's the links for the workshop:
http://forommm.medialab-prado.es/viewtopic.php?id=213

so what we are doing is trying to create logic switches out of the PH
conductivity taken from fruit juices - we are trying to develop a system
that uses different juices (pHs like from lemons or from peas) that yield
a 0 or a 1 - this signal is taken from an electrode and transferred to a
resistor - then to a microcontroller to a program (processing) that gives feedback.
Have you worked on anything like this before or do you know anyone that has?
Just looking for ideas to bring into the project which will eventually be distributed
as an instructible.
Right now we are experimenting with the probe which is a conductivity cell - but
we are making it from scratch so... thinking of materials/vessels...ect. Also need
a way to add or take away liquids - was thinking some kind of drainage or siphoning
device. If you have any thoughts we would really appreciate it!
</snip>

Cheers,
Jason Bobe

---------- Forwarded message ----------
From: Jason Bobe <jaso...@gmail.com>
Date: Fri, Feb 6, 2009 at 1:46 PM
Subject: Re: ideas for organic computer? (based on pH of fruit juices)

Sounds great Kelly.  It appears this fruit computer thread never made it to the google group, so I'm sending it along now.  I'm sure you'll get some valuable feedback.

Cheers,
Jason Bobe

On Fri, Feb 6, 2009 at 11:07 AM, Kelly Andres <kelly....@gmail.com> wrote:

Hey Mac!
Thanks, this is great info. We have had a bit of trouble with the electrode we made so we are going
with a PH probe hack jsut to get things running - it seems that the electrode we built was acting like a
little battery - so measuring the sensitivity we needed to get a 0 and 1 from the addition/subtraction of
juices became difficult... well it did not work.
Will keep you all updated! And the project wiki is here - that is where the info on the project will be kept.
http://wiki.medialab-prado.es/index.php/Fruit_Computer_Laboratory

kelly

On Fri, Feb 6, 2009 at 12:08 AM, Mackenzie Cowell <maco...@gmail.com> wrote:

Your project sounds interesting, Kelly.  In the iGEM crowd (igem.org), organic computer projects tend to be implemented with protein - dna interactions inside living cells.  Most of the the projects don't work.  It's refreshing to hear of a completely different approach.

That said, Kay Aull is about halfway to the point of having a single bit molecular flip-flop in e. coli: http://io9.com/5049788/making-a-biological-counter.  I'm CCing her on this thread in case she has any suggestions.  Perhaps you could eventually develop a way to signal from your system to kay's cells. 

Additionally, I am sending this message to the DIYbio google group for their feedback.  There are a lot of smart people over there.

For your liquid handling needs, I have a couple of suggestions:

- small plastic syringes (10-50 mL) with the plunger connected to some kind of linear actuator (car doors have cheapo solenoids in them) can be used to push and pull fixed amounts of liquid around.

- alternatively, I think I have seen a kind of linear peristaltic pump based on a single solenoid oriented perpendicularly to some flexible tubing, such that on extension, the tubing was compressed.  Apparently most peristaltic pumps use rotational motion: http://en.wikipedia.org/wiki/Peristaltic_pump.

Good Luck!

Mac
(always been a fan of processing, btw)

On Tue, Feb 3, 2009 at 11:08 AM, Jason Bobe <jaso...@gmail.com> wrote:

Hi Kelly -

Nice to hear from you.  The project sounds fascinating.  I'm CCing Mac and Jason Morrison, both of whom may have some experience that could be useful.

I think the best way to get feedback would be to post your message below (or one similar) to the DIYbio google group.  There are nearly 600 people on the list now and many of these folks may be able to provide technical counsel for you.

http://groups.google.com/group/diybio

If you would prefer me to post a message on your behalf, I'm happy to do that too.  Let me know.

Thanks!
Jason


Jason Bobe
Director of Community
Personal Genome Project
Harvard Medical School
www.personalgenomes.org

On Tue, Feb 3, 2009 at 10:57 AM, Kelly Andres <kelly....@gmail.com> wrote:

hi Jason
We met at MIT during the symposium in the lab where you guys gave
your diy DNA extraction and i was wondering if you have any ideas for
a project i am collaborating on at the garage science workshop at media
lab prado. Here's the links for the workshop:
http://forommm.medialab-prado.es/viewtopic.php?id=213

so what we are doing is trying to create logic switches out of the PH
conductivity taken from fruit juices - we are trying to develop a system
that uses different juices (pHs like from lemons or from peas) that yield
a 0 or a 1 - this signal is taken from an electrode and transferred to a
resistor - then to a microcontroller to a program (processing) that gives feedback.
Have you worked on anything like this before or do you know anyone that has?
Just looking for ideas to bring into the project which will eventually be distributed
as an instructible.
Right now we are experimenting with the probe which is a conductivity cell - but
we are making it from scratch so... thinking of materials/vessels...ect. Also need
a way to add or take away liquids - was thinking some kind of drainage or siphoning
device. If you have any thoughts we would really appreciate it!
thanks
kelly

[JonathanCline]

On Feb 6, 12:51 pm, Jason Bobe <jasonb...@gmail.com> wrote:
> Kelly Andres is working on a fruit computer laboratory

> > so what we are doing is trying to create logic switches out of the PH
> > conductivity taken from fruit juices

> > Right now we are experimenting with the probe which is a conductivity cell

pH is a stable signal. How does it change to make Hi and Lo? Review
logic gates:

http://en.wikipedia.org/wiki/Logic_gate

The output needs to change based on something. Are you using heat?
Conceivably and completely hypothetically an Apple gate would take two
inputs: apple juice and heat, and give one output: pH (measured by
the conductivity probe). If Apple juice exists and heat exists (to
warm the Apple juice), then output is Lo pH. If Apple juice exists
and heat does not exist, then output is Hi pH. If Apple juice does
not exist (or water used instead), then the heat input is "don't care"
and the output is Lo. However: if the Apple juice is later applied
to the same input, it has to return to it's normal state (i.e.
regardless of heat; non-denatured?), which is where biology/chemistry
has the problem, right? i.e. reactions are usually one-way. So it
would seem the real problem is in making the biology/chemistry to be
reversible reactions. If you have "perfectly" reversible reactions
(meaning: pH can be Hi then Lo then Hi then Lo, in response to some
external something), then it can be done, and the only limit is on
physical size. However nothing is perfect, so there's bound to be
some kind of loss somewhere (in silicon, this loss is electrical
current which creates heat). So something hints that the Apple juice
won't return to normal after the heat (or light, or whatever) is
repeatedly applied & removed.

At least that's my untrained view.. I might be completely wrong. Hm,
wikipedia to the rescue: http://en.wikipedia.org/wiki/Chemical_computer

If the project is pumping different liquid around as inputs to a
microcontroller, then it sounds more like using-fruit-as-a-keyboard.
i.e. the presence of pH is like pressing an On-Off switch, like a key
on a keyboard, or like a rube goldberg device (lots of triggers). The
actual computing would be in the silicon.

Unless you're building an analog computer, which might sum the pH's to
yield different results. Then you're not looking for Hi and Lo,
because analog computation can be different kinds of math (like: lo,
med, hi) as long as everything fits into those levels. In which case,
Apple juice would always indicate 1, orange juice might be 2, kiwi
might be 3, and water might be zero, giving you a base-4 system. Hey,
that's a nibble when using fruits, isn't it. Ok, bad pun. Anyway,
that's more like an abacus (different weight systems) than a digital/
binary computer where everything is either 1 or 0.
http://en.wikipedia.org/wiki/Analog_computer

Either way, it could be done. The media lab has made stranger stuff,
and NASA sent guys to the moon using slide rules, so it's possible.
And sticky. BTW, you might try using diet coke. It would clean out
the works pretty well, and might make a good zero. Diet coke
dissolves everything..


## Jonathan Cline
## jcl...@ieee.org
## Mobile: +1-805-617-0223
########################

[Kay Aull]

I suppose you could use the tubes of fruit juice as a kind of memory
device, with the pH being a readout of what is in each tube. You'd
still need some sort of external logic, like:

if (Tube1.pH == LEMON.pH):
replace Tube2 with CARROT

I'm making that up...but you get the idea. There needs to be some
rules to compute by, and some read/write system. Otherwise you've got
a fruit juice stand, not a computer. =)

If you haven't already, you might read up on Turing machines - you
could make one using the fruit juice tubes as "memory tape".

[Len Sassaman]

I've been working on the idea of a bacteria-based computer, using a
different approach to producing and storing binary states than pH
conductivity -- that's quite clever! -- and would love to talk to Kelly
about this sort of thing.


Cheers,

Len

Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm

[Nick Taylor]

Hmmm... reminds me of this project that my dad got his kids to do at the school he taught at up the river in the 60s

http://www.freewallpaper.co.nz/images/1920x1200/Jerusalem.jpg

They made abacuses out of Life-Savers - little donut-shaped fruit sweets. Worked very well... for a while. 

I'm not sure if that counts as fruit exactly. Wasn't one of the early computers a Cabbage Engine or something?

[Bryan Bishop]

On Fri, Feb 6, 2009 at 7:20 PM, Len Sassaman wrote:
> I've been working on the idea of a bacteria-based computer, using a
> different approach to producing and storing binary states than pH
> conductivity -- that's quite clever! -- and would love to talk to Kelly
> about this sort of thing.

Over the summer I was exploring some work on DNA logic. It's slow (and
the logic operations remain incomplete). It's not going to compete
with silicon computation. The classic example is the ribozyme
tic-tac-toe game that takes 30 minutes per move for the decision tree
to propagate. There might be something else that it could do though-
maybe at the intersection of fabrication and computation. I guess this
is when you cite Erik Winfree and Paul re: self-assembling DNA
patterned stuffs. As for bacterial computation, maybe there's some
speed you can pick up, somehow, somewhere?

- Bryan
http://heybryan.org/
1 512 203 0507

[ben lipkowitz]

On Sat, 7 Feb 2009, Bryan Bishop wrote:

> Over the summer I was exploring some work on DNA logic. It's slow

<snip>

> As for bacterial computation, maybe there's some
> speed you can pick up, somehow, somewhere?

I always thought that the rod logic described in The Diamond Age would be
a good match for protein based machinery.

http://cse.stanford.edu/classes/sophomore-college/projects-98/babbage/ana-mech.htm

now if that doesn't look like a protein complex, i don't know what does.
these CPU protein complexes could contain an RNA polymerase enzyme in
order to "network" with each other by sending short RNA "data packet"
molecules ending with the unique destination address. it would be slow,
but would enable a vast number of CPU's to operate in parallel in the
same cell.

of course it begs the question: how do you build it?

[phillyj]

a pH computer would be very hard to maintain, I think. The pH would
have to stand constant and probably minute pH changes would affect
calculations. The set up would be large scale until someone figures a
microscale pH computer.

The next dimension of computing is Quantum Computing.

http://en.wikipedia.org/wiki/Quantum_computer
http://computer.howstuffworks.com/quantum-computer.htm

Back to organic computing, possibly something mimicking the membrane
potential of neurons.

On Feb 7, 10:48 pm, ben lipkowitz <f...@sdf.lonestar.org> wrote:
> On Sat, 7 Feb 2009, Bryan Bishop wrote:
> > Over the summer I was exploring some work on DNA logic. It's slow
> <snip>
> > As for bacterial computation, maybe there's some
> > speed you can pick up, somehow, somewhere?
>
> I always thought that the rod logic described in The Diamond Age would be
> a good match for protein based machinery.
>

> http://cse.stanford.edu/classes/sophomore-college/projects-98/babbage...

[Len Sassaman]

On Sun, 8 Feb 2009, phillyj wrote:

>
> a pH computer would be very hard to maintain, I think. The pH would
> have to stand constant and probably minute pH changes would affect
> calculations. The set up would be large scale until someone figures a
> microscale pH computer.

I agree with this. However, pH is one of many biological channels that can
be used for transmitting information and storing state (for a finite
amount of time).

> The next dimension of computing is Quantum Computing.

There are many "next dimensions" in computing. Personally, I think quantum
computing is a small advance compared to where we are today; DNA computing
is much more promising to me.

http://en.wikipedia.org/wiki/DNA_computing

> Back to organic computing, possibly something mimicking the membrane
> potential of neurons.

I've been looking at the ion channel pathways, and using bacteria as
transisters (with Meredith).

There's a *lot* of potential (no pun intended) future direction here.

--Len.

Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm

[Bryan Bishop]

On Sun, Feb 8, 2009 at 2:41 PM, Len Sassaman wrote:
>> Back to organic computing, possibly something mimicking the membrane
>> potential of neurons.
>
> I've been looking at the ion channel pathways, and using bacteria as
> transisters (with Meredith).

There used to be a guy running around on the internet that had written
a compiler, or rather the grammar to generate the compiler, that took
ANSI C programs and translated it into artificial neural network
layouts. These networks then performed the computations. Combined with
the long history of growing neurons on HMDI silicon surfaces, and
groups making neurons grow on circuit pathways via coating the paths
with neural growth factors, there might be something worth exploring
there.

[Jeswin John]

Bryan, Do you remember a while back about some scientists using snail neurons in their computing research?
This is a similar article: http://www.livescience.com/health/060327_neuro_chips.html

Does anyone know how that is going, any updates?

I think DNA computing would be of great importance in the artificial creation of chemicals and organic compounds via natural pathways. Am I correct?

--
*-----------------------------------------------------------*
Join the Revolution

http://diybio.org/
http://homebrewbioscience.blogspot.com/
*------------------------------------------------------------*

[Gene Hacker]

How about making a computer using John Conways game of life in a
Belousov-Zhabotinsky Medium. (http://en.wikipedia.org/wiki/Belousov-
Zhabotinsky_reaction)

It's recently been shown that one can make glider guns in such a
system, so why not try making a simple computer? http://arxivblog.com/?p=1183

John Conway's game of life has similar computational ability to that
of a Universal Turing Machine. One might be able to make a very
powerful computer using just a jar of chemicals!

One might also try something like this:
http://vimeo.com/3124876

[Nick Taylor]

Gene Hacker wrote:
> How about making a computer using John Conways game of life in a
> Belousov-Zhabotinsky Medium. (http://en.wikipedia.org/wiki/Belousov-
> Zhabotinsky_reaction)

There's an excellent cellular automata site here:

http://www.collidoscope.com/modernca/welcome.html

n