Fwd: BioPrinter won 1st Prize in Instructables Design Competition!

[Bryan Bishop]

From: Patrik D'haeseleer <pat...@gmail.com>
Date: Thu, Feb 14, 2013 at 12:37 PM
Subject: [biocurious] BioPrinter won 1st Prize in Instructables Design Competition!
To: biocurious <biocu...@googlegroups.com>, BioCurious Printer Hacking <biocurious-pr...@googlegroups.com>


Woohoo - Our BioPrinter Instructable won a (shared) 1st Prize in the Instructables Design Competition, out of 917 competitors!

This is by far our biggest prize to date - a 13" 2.9GHz Macbook Pro, which I think is something like a $1500 value. I think this will now become the official BioCurious Laptop. Unless people think we are already well served with the tiny laptop at the lab, in which case this could become one of the prizes for our own internal Instructables Incentive Program...

Thanks a bunch to everyone who helped make this possible, and especially to Paul Miller in the BioPrinter team - without his clean design for the bioprinter, there's no way we could have snuck our way into a *design* competition like this.

So - let's get cracking on some more Instructables, people! I especially have my eyes on the UP! 3D Printer contest ending April 1st, which has 10 3D printers to give away.

Patrik

--
--
You received this message because you are subscribed to the Google
Groups "BioCurious" group.
To post to this group, send email to biocu...@googlegroups.com
To unsubscribe from this group, send email to
biocurious+...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/biocurious?hl=en
 
---
You received this message because you are subscribed to the Google Groups "BioCurious" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biocurious+...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

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

[Humpel]

Congratulations!  

But rather sell the apple crap and buy a few real laptops with the money xD

[Cathal Garvey]

For that price you could set up a neat little garage lab.. :)

On 16/02/13 12:48, Humpel wrote:
>
> Congratulations!
>
> But rather sell the apple crap and buy a few real laptops with the money xD
>

> --
> -- You received this message because you are subscribed to the Google

> Groups DIYbio group. To post to this group, send email to
> diy...@googlegroups.com. To unsubscribe from this group, send email to
> diybio+un...@googlegroups.com. For more options, visit this group
> at https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org

> ---
> You received this message because you are subscribed to the Google

> Groups "DIYbio" group.

> To unsubscribe from this group and stop receiving emails from it, send

> an email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio?hl=en.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/diybio/-/vK3_m5f2k8UJ.

[Jonathan Cline]

From: Patrik D'haeseleer <pat...@gmail.com>

Woohoo - Our BioPrinter Instructable won a (shared) 1st Prize in the Instructables Design Competition, out of 917 competitors!

Congrats and very cool platform.

Though not sure why go the X-Y rectangular axis system at all, and why inkjet technology is needed as the liquid dispenser (which brings with it significant challenges, i.e. clogging, pore size, etc).  You had the working compact disc drive mechanisms with their working polar-axis control system (linear actuator for 'r', drive spindle for theta; perhaps replace the spindle motor with stepper for easy positioning without requiring an encoder wheel).  Petri dishes are already circular so would integrate well directly with the existing polar (i.e. rotational) disc drive mechanism, i.e. spin the dish and address points in the dish via sector (angle) and track (radius).  The liquid dispensing could be done by very small diameter tubing run through a peristaltic pump from a large well/beaker source which dispenses drops as needed from above, positioned via the linear actuator -- such pumps are very accurate, to uL's or nL's.  Add multiple lines from multiple pumps for dispensing multiple reagents, arrange these in a row on the radial axis.

X-Y axis as seen in typical lab robotics is needed to mimic the positioning of humans with minds trained for cartesian addresses and grasping by the shape of human hands which have thumbs.  Motors on robots do not have good thumbs, they have spinning shafts and work better with polar axis.  Which is why hard drives, compact discs, floppies, are all circular, rather than square.  When designing new machines and automating protocols, fit the protocol/equip to the new machine not based on human limitations for best results, rather than build a machine which for traditional/limiting human-style equipment.

If there are points to the contrary I'd love to hear them.


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

[Nathan McCorkle]

On Fri, Mar 1, 2013 at 12:58 AM, Jonathan Cline <jnc...@gmail.com> wrote:
>
>
>>
>> From: Patrik D'haeseleer <pat...@gmail.com>
>>
>> Woohoo - Our BioPrinter Instructable won a (shared) 1st Prize in the
>> Instructables Design Competition, out of 917 competitors!
>
>
>
> Congrats and very cool platform.
>
> Though not sure why go the X-Y rectangular axis system at all, and why
> inkjet technology is needed as the liquid dispenser (which brings with it
> significant challenges, i.e. clogging, pore size, etc). You had the working
> compact disc drive mechanisms with their working polar-axis control system
> (linear actuator for 'r', drive spindle for theta; perhaps replace the
> spindle motor with stepper for easy positioning without requiring an encoder
> wheel). Petri dishes are already circular so would integrate well directly
> with the existing polar (i.e. rotational) disc drive mechanism, i.e. spin
> the dish and address points in the dish via sector (angle) and track
> (radius). The liquid dispensing could be done by very small diameter tubing
> run through a peristaltic pump from a large well/beaker source which
> dispenses drops as needed from above, positioned via the linear actuator --
> such pumps are very accurate, to uL's or nL's.

What would you propose for a pulse-free pumping solution in the uL/nL
range? I know a pump can handle those volumes, but what kind of tip
would let go of uL or nL droplets? The printer cartridge ejects pL
droplets. I've also heard of using high-frequency waves to eject
controlled size droplets.

> Add multiple lines from
> multiple pumps for dispensing multiple reagents, arrange these in a row on
> the radial axis.
>
> X-Y axis as seen in typical lab robotics is needed to mimic the positioning
> of humans with minds trained for cartesian addresses and grasping by the
> shape of human hands which have thumbs. Motors on robots do not have good
> thumbs, they have spinning shafts and work better with polar axis. Which is
> why hard drives, compact discs, floppies, are all circular, rather than
> square. When designing new machines and automating protocols, fit the
> protocol/equip to the new machine not based on human limitations for best
> results, rather than build a machine which for traditional/limiting
> human-style equipment.
>
> If there are points to the contrary I'd love to hear them.
>
>
> ## Jonathan Cline
> ## jcl...@ieee.org
> ## Mobile: +1-805-617-0223
> ########################
>

> --
> -- You received this message because you are subscribed to the Google Groups

> DIYbio group. To post to this group, send email to diy...@googlegroups.com.

> To unsubscribe from this group, send email to

> diybio+un...@googlegroups.com. For more options, visit this group at
> https://groups.google.com/d/forum/diybio?hl=en
> Learn more at www.diybio.org

> ---
> You received this message because you are subscribed to the Google Groups

> "DIYbio" group.

> To unsubscribe from this group and stop receiving emails from it, send an

> email to diybio+un...@googlegroups.com.
> To post to this group, send email to diy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/diybio?hl=en.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/diybio/-/fYFrR2dePdYJ.

>
> For more options, visit https://groups.google.com/groups/opt_out.
>
>



--

-Nathan

[Patrik D'haeseleer]

Sounds like an interesting idea, and I look forward to seeing your version! ;-)

Mainly, we went with this approach because both the XY platform and the inkjet printing had already been demonstrated, and seemed easiest to combine. I think modifying an actual CD drive for polar printing onto a standard Petri dish would be very difficult. And if you have to engineer it from scratch anyway, putting together a square Cartesian system is no harder than doing a polar system.

As for the inkjet print head - it definitely works, and this method is in use in a few academic labs, but we are indeed planning to switch to a syringe pump design for increased flexibility.

Patrik

[John Griessen]

On 03/01/2013 02:58 AM, Jonathan Cline wrote:
> The liquid dispensing could be done by very small diameter tubing run through a peristaltic pump from a large well/beaker source
> which dispenses drops as needed from above

Like Nathan said. Are you proposing contact dispensing to transfer liquid that otherwise would be trapped by surface tension?
The jet part of ink jet overcomes surface tension and gives you contactless dispensing.

[John Griessen]

On 03/01/2013 06:24 AM, Patrik D'haeseleer wrote:
> I think modifying an actual CD drive for polar printing onto a standard Petri dish would be very difficult.

He was pointing out that it is easy. The linear drive from the CDROM is one axis -- you locate it
from center to radius MaxR, then install a stepper motor with stout shaft under point called center,
then attache a petri dish holder to the stout shaft. Voila -- petri dish coordinates.

And if you have to
> engineer it from scratch anyway, putting together a square Cartesian system is no harder than doing a polar system.

X-Y seems harder than adding a disk to a stepper shaft to me...while alternatively polar coords simplifies your
code considerably for access to the full area of petri dishes.

[Patrik D'haeseleer]

On Friday, March 1, 2013 11:45:38 AM UTC-8, John Griessen wrote:

On 03/01/2013 06:24 AM, Patrik D'haeseleer wrote:
> I think modifying an actual CD drive for polar printing onto a standard Petri dish would be very difficult.

He was pointing out that it is easy.  The linear drive from the CDROM is one axis -- you locate it
from center to radius MaxR, then install a stepper motor with stout shaft under point called center,
then attache a petri dish holder to the stout shaft.  Voila -- petri dish coordinates.

I'm not saying it's impossible, but it would definitely be harder than the way we did it. For one, you'd have to take the whole thing out of the CD drive chassis anyway, because there's no way you'd be able to fit this entire contraption "in situ", as it were. Then you'd have to engineer a turntable from scratch, because you should probabaly use a stepper motor instead of the spindle motor in the drive. You'll also need some way to keep the Petri dish in place as it's spinning, because I don't think you'd want to use the same clamping mechanism used to hold a CD.

In the end, it comes down to one linear actuator + a Petri dish turn table built from scratch, versus two linear actuators. Disassembling and controlling two linear actuators is no harder than one (which you'd need for polar coordinates anyway), so the second solution is definitely easier to build. Not to mention the issues you may have with variable resolution in polar coordinates, having to do coordinate transformations for everything, the ready availability of cartesian-based stepper controllers and software from the 3D printing field, etc.

I really would love to see someone build a bioprinter that uses polar coordinates. But so far I haven't heard any arguments to make me think that approach would be any easier or better than what we did.

If someone feels inspired - go for it though!

Patrik

[Jonathan Cline]

On Fri, Mar 1, 2013 at 4:18 AM, Nathan McCorkle <nmz...@gmail.com> wrote:

What would you propose for a pulse-free pumping solution in the uL/nL
range? I know a pump can handle those volumes, but what kind of tip
would let go of uL or nL droplets? The printer cartridge ejects pL
droplets.

First explain why you need nL or pL droplets.  Yeah I know it's all the rage at MIT to constantly make everything smaller for apparently little purpose other than to claim it's novel and get a mention in Wired.   From the instructable directions, there's paragraphs explaining that using a special inkjet head with fatter nozzles is needed to overcome problems anyway, i.e. 'The problem with inkjets is that they're resolution is too *high* and we need lower resolution' (i.e. fat pipes).   For a petri dish pL or nL is not needed.. for microarray they are not needed..  Even if you're going to try microfluidics with pL volumes you'll run into other problems (how about evaporation?) so as mentioned before, perhaps the best hack right now is milli-fluidics not micro.    For microliters the commercial comparison is Biotek microflo (likely others, that's just one I evaulated) and it uses plastic tips and tubes to dispense 0.5 uL - 3,000 uL properly.

Ref http://www.biotek.com/products/liquid_handling/multiflo_microplate_dispenser.html

Volume Range

500 nL - 3,000 µL/well Selectable in 1 µL increments

--

[Patrik D'haeseleer]

Another issue in bioprinting is that you typically don't really *want* to position each cell individually. It's much more efficient to print a clump of cells at a time, and allow them to self-organize.

Printing larger droplets also means you can print larger volumes faster. When printing living 3D structures, it's a big challenge to make sure the cells on the bottom are still alive by the time you finish printing the cells on top. Inkjet is fast, but the total accumulated volume is quite small.

Patrik

[Jonathan Cline]

On Fri, Mar 1, 2013 at 2:29 PM, Patrik D'haeseleer <pat...@gmail.com> wrote:

I really would love to see someone build a bioprinter that uses polar coordinates. But so far I haven't heard any arguments to make me think that approach would be any easier or better than what we did.

Automation flow.  The next stage after your liquid handling is manual removal to an incubator.  Then removal from the incubator to the imaging station.  Then pick-and-place at the selection station.  Then sample purification.  Then culturing.  etc.  How many of these stages can be combined into a 1-button-push "make me bread" machine and how many of these stages are unique pieces of equipment.  When moving a sample dish, pick-and-place is easier if any orientation of the sample dish is allowed (hence, sample dish is circular,   symmetric).  Imaging/Measurement works great if the sample dish is rotated past a scanning head (hence, sample dish is circular and rotates).  Etc.  The cartesian/human/traditional method is to slap the sample plate on a large scanner.  Hmm.  Well wouldn't a microelectronics solution point to a laser-disc type approach with a high precision micro sensor scanned across the sample, rather than increasing the physical size of the sensor into big flat bed?

I've asked a dozen lab automation experts these same types of questions and received blank looks so either it is a really dumb idea or it's so obvious that no one has thought of it yet.  hah!   Actually it's because the current state of the industry is a mix of human hands and machine.  The human hands require rectangles.

--

[Patrik D'haeseleer]

That's a very different application though. Our bioprinter is not intended to be a general lab robot or liquid handling tool. It's intended to print (and eventually 3D print) biological materials onto a surface.

You've got the wrong BioCurious robot in mind ;-)

Patrik

[John Griessen]

On 03/01/2013 05:53 PM, Jonathan Cline wrote:

> On Fri, Mar 1, 2013 at 2:29 PM, Patrik D'haeseleer <pat...@gmail.com <mailto:pat...@gmail.com>> wrote:
> I really would love to see someone build a bioprinter that uses polar coordinates. But so far I haven't heard any arguments to
> make me think that approach would be any easier or better than what we did.
>
> Automation flow.

.
.
.
.

> I've asked a dozen lab automation experts these same types of questions and received blank looks so either it is a really dumb
> idea or it's so obvious that no one has thought of it yet. hah! Actually it's because the current state of the industry is a
> mix of human hands and machine. The human hands require rectangles.

Maybe there is a breakthrough on the verge of happening as people finally give up
doing so much with warm bodies carrying and positioning things, (inaccurately, and with dirt),
and start doing more programming of networks of liquid/vial/plate/capsule/capillary/slide/array/microarray machines.

I think it is going to be ethernet connected machines, how about you?

[Jonathan Cline]

I'm not sure what you mean by "contact dispensing" but the peristaltic pump approach can eject droplets into the air at a pretty good velocity, centimeters away from a well.  i.e. either droplets or a stream of liquid can fly many centimeters beyond the end of the tube.   Probably depends on the tip shape but I doubt it's rocket science either.  Peristaltic pumps with the right tubing can generate good amount of pressure.  The design of the 'wheel' of the pump and the gearing/type of the motor determines the steadiness of the stream (i.e. does the pressure of the stream oscillate or is it steady pressure).   Stepper motor for example creates more variation in pressure since the motor "locks" to the magnet at each step.   When I asked the lead designers at a couple different pump companies why they didn't use DC motors for smooth dispensing they said: "Well our engineers didn't want to bother with PID controllers and it doesn't really matter anyway [i.e. doesn't affect results]".   Naturally using a dc motor with PID algorithm would be part-wise $$ cheaper (and electrically less power) than a stepper. ..that's a whole other tangent..


If I remember right, inkjet heads have good velocity & accuracy many centimeters away too.  I think there was a graffiti artist who was using an inkjet head to spray ink a foot away from the platform surface without any slop.

If the goal is to make a 3D bio-printer then the next question is whether the printer should move the print head vertically or move the platform vertically.  I think repraps faced the same conundrum...

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

[Jonathan Cline]

On Saturday, March 2, 2013 10:52:40 AM UTC-8, John Griessen wrote:

start doing more programming of networks of liquid/vial/plate/capsule/capillary/slide/array/microarray machines.

I think it is going to be ethernet connected machines, how about you?

I think the future is in USB peripherals really, which are then connected to a larger machine (or larger embedded computer, like Raspberry Pi) which in turn is on the network (likely via wifi).  The lab devices present themselves as serial ports and are controlled with simple text commands (aka virtual COM ports for windows people), optionally using USB power where possible.   This keeps the individual lab devices relatively simple.  And a lab could mix-and-match only those devices they need.  USB cables and hubs have come down in price to be near free, and USB extender cables eliminate cable length problems ("active" extension cables with built-in hub to stay within the USB signal limitation). I've never been much of a proponent for embedded network software or embedded network ports -- even after building these devices -- because it adds a big chunk of complexity to an otherwise simple instrument (for example, most bio instruments are really only controlling a couple different things, like sensor on/off or single data acquisition measurement, etc).  USB 2.0 has plenty of bandwidth for higher bandwidth devices like imagers.  Most 90's era devices if they included automation/computer interfaces used industrial-strength serial ports and simple text commands, even complex robots used this method, see the Tecan command set for example (google "perl tecan yaml").  These ports can be replaced with simple USB without limits on the number of physical ports, which was such a problem with old serial devices.  Simple text commands allows any computer software, in any language, to control the machinery (although even Tecan put security in the middle which required paying Tecan $$$ to remove).  The lab automation industry itself has been pushing for ethernet-enabled devices because it means they can sell their proprietary software into the device maker's products and then turn around and sell their proprietary software into the PC application products too (on the similar front, they're trying this with USB device drivers).

Additionally this means there is an intelligent networked device (or PC) which controls these USB lab peripherals.  The PC has virtually unlimited resources so it can act as either a protocol server or web app server or some enterprise-class database-integrated behemoth server (larger scale lab automation absolutely needs to be tied into the inventory database, for pulling the right reagents from the right locations, or checking if needed equipment is online and available, posting results data, etc etc).  Tablets/handsets connect to the devices through the PC web-app server which is the typical model for communication now.   (Even when IPv6 arrives and every device could theoretically have direct internet addressing, the model will still likely be client->server->device, not client->device.)

Lab devices don't move around, so there's not much need for wireless, though it would be "cool".  It's also much tougher to sell a niche wireless device (certification requirements, etc) than to sell a wired device.
 
Years from now the solution will be wireless power (which still presents itself as a virtual COM port for controlling the device, through the wireless power's communication link ;-D ) -- though wireless power is not cheap enough or ubiquitous enough yet (google "wireless blender").

If going ethernet then the natural trend would be to try power-over-ethernet too, to eliminate the power cords, but that also is not cheap enough or ubiquitous enough.  When I say ubiquitous, the measurement is seeing a huge box at Fry's with a big sticker that says "$1.99 surplus".  The world prefers USB to POE.

I think the bigger trend is in killing the silly membrane/touch keypad and LCD/VGA display from the instruments themselves.  That's 50% of the manufacturing cost (or more) right there.  Move the user interface onto the PC or tablet or smart phone and let the lab device just be completely enclosed.  The devices themselves should have no more than 2 push buttons which covers the operations of: power toggle and "go/stop"; and no more than 2 LED's to indicate operational status.  That's old-school getter-done design.

[Bryan Bishop]

On Sat, Mar 2, 2013 at 10:01 PM, Jonathan Cline wrote:
> This keeps the individual lab devices relatively simple. And a lab could
> mix-and-match only those devices they need.

I think that the best case scenario is reducing the amount of "crap"
between servos/steppers, other actuators, and a usable not-locked-down
computer. Once you get control up to a general computer, it's not as
problematic to go for ethernet or bluetooth or gsm or whatever the
flavor of the week is.

One of the issues that keeps coming up when people talk with me about
lab automation is how to organize the locus of control of some system.
A propellor? arduino? raspberry pi? For most lab tasks, there's no
need for sub-microsecond timing, so you don't even need rtlinux and
could get away with a general-purpose computer. This has some
advantages like packages and you don't have to worry about manually
managing binaries on your ROM or whatever, except for the user-space
program that's running. It used to be that computers cost too much,
but computing and memory are very low cost these days, so I don't see
why not.. plus you can take advantage of programming time spent across
multiple systems, since most of these hardware projects turn out to be
very similar ("rotate 1/2pi, engage girder").

Jonathan, I think your experience as maintainer of Robotics::Tecan
uniquely positions you to have some informed opinions about how to do
libraries or development for lab hardware, and I think the community
would benefit from you ranting ("documenting") more. For instance, it
would be great if we could somehow convince, just as an example,
Dietrich at BioCurious, to use Robotics::Tecan (or something similar)
and maybe even something approaching a toolchain for compiling or
deploying ROMs... big dreams.