Re: [DIYbio] OpenTrons BioBot: Open-Source DIY Liquid Handling Robot!

Thanks for this! Great ideas.

Camera: We are on 100% the same page here. We are planning on incorporating an Rpi camera for auto calibration and job-monitoring for the BioBot One that we will be Kickstarting in November. We're currently finishing up the physical design for manufacture, and  porting the software from Android/TinyG to Rpi/Smoothieboard. We should talk more off-list about developing the computer vision software going forward! 

Enclosure: this is certainly an awesome idea, and we have been playing with it ourselves. It adds a ton of cost however, and would require a lot of testing to get right, so we've decided to hold off and include this in later models of the 'bot. But that is no reason to stop you from experimenting! Would love to hear about anything you discover. Out of curiosity, what are you hoping to use the 'bot for that you'd want environmental controls like this? 

Automatic Tool Change: This is something we actually have made some designs for, but it is also a feature that is going to have to wait for the next version. But to compensate for not having auto tool changes on the BioBot One, we are including three active heads so you can have two differently sized micropipettes and a test-tube gripper all at the same time -- hopefully this makes it flexible enough to be useful. 

AND-- 

Open Source Protocol Scripting and Sharing: We are including compatibility with PRPR -- an open-source liquid handling robot scripting language alternative to the proprietary Tecan EvoScript or similar -- in all of our labtools. This means that you can write code to do biotech protocols on your bench top. We are developing a community hub for sharing and creating these scripts as well as a node.js API/Library. We think that enabling crowdsourced optimization and quick download and run capabilities could be really awesome for empowering biodevelopers and improving scientific reproducibility -- but this is just getting started, and it would be amazing to hear anyone's thoughts on sharing scripted protocols like this! 

THANKS!!!!!!

Will 

On Sunday, August 3, 2014 4:47:34 PM UTC+8, Michael Shamberger wrote:

I have been thinking about some other design changes that would be good for a bio bot.

- Enclosure to maintain specific environment.   This allows it to be used as an incubator.  Cooling would allow it to be used for printing microarrays.  The enclosure would be made of correx and could be paired with a HEPA filter.

See http://users.ox.ac.uk/~atdgroup/technicalnotes/TE300%20incubator%20enclosure.pdf for an initial design that is compatible with the current opentrons.

- Computer vision for self calibration and "Pick and Place".  I am planning on using a downfacing Raspberry PI camera.  I think I will start with the vision library made my guys working on the Firepick delta (www.firepick.org) which is a pick and place machine.  See the code at github https://github.com/firepick1/FireSight.  It is a layer on top of opencv and runs on Raspberry PI.  The computer vision can recognize the plates being used for experiments.

- Automatic tool change needs more thought

I haven't started my build yet because I have been investigating these design changes and will make CAD drawings before physical assembly.