OpenTrons: Open Source Robotic Liquid Handling
opentrons.com
opentrons.com
Why not use an electronic pipet that does all of that for you? I remember using a digital pipettor years ago that was already programmable for just this type of application (eppendorf maybe?).
That's really the issue that needs to be front and center - how reproducible is this machine. For any lab that could use one of these, that is always the primary concern. What I'd like to see is something like a dye test where you diluted a dye into each well on a 96 well plate and then used a plate reader to determine how precise you are.
The second consideration is speed. How fast can this possibly be? I've used a liquid handling robot before that "only" had a 4-channel head, and it was slow enough that I eventually gave up and set the plates up by hand with an 8-channel pipet. Precision is the main concern, but speed is a close second. If you want to replace manual pipetting, you better be faster than a grad student.
And if a lab is able to use automation, the P1000 in the front page video is not going to be what they need. There are few applications where you'll want to pipet 200-1000ul in an automated manner - there are many more applications for 1-10ul. So I'd like to see more information on what pipettors are supported, and how do you change them?
As someone below mentioned - there is just too much missing information on the page. I had to resize my browser just to see the full video, which answered a couple of questions, but not all.
How many labs are currently using your machine, and for how many are these just toys to play with vs production machines? Meaning - if a lab is setup to properly test the machine, they probably have a few other liquid handlers in place. How many of their protocols can they transfer to the OT-One? And would any of them choose this over a more advanced machine?
As I said below, we have shipped 51 robots so far and most are in daily use by the academic research labs and biotech companies that bought them. Serial dilutions, mother daughter plating, ELISA, PCR prep, ligations/digestions, heat shock transformations, and soon Gibson assembly are the applications most widely run on the OT-One.
Yes we do depress the top of the manual pipette, just like you would do by hand. Do you trust your manual pipetting to be reproducible? Thats exactly the same thing that is happening here. Agreed we need a dye test to show reproducibility/precision/accuracy, and it is currently in the works (you can find a scale test showing accuracy/precision on our blog, blog.opentrons.com). But from running the machine for 6 months and getting good biological data on it, I can tell you that it reliably hits the desired volume when correctly calibrated. And it can hit different volumes each transfer w/o setting the pipette, it just presses the plunger down the desired amount (say you want 100ul using a p200, it pushes down 50%). The electronic pipettes you are talking about cost more than our whole robot.
Our current users disagree with you that we need to be faster than a grad student, thought when turned up to full speed it is faster (but needs to be bolted to the bench when you turn it up that fast). In fact, the grad students that use the machine love it because they can go do other things and not worry about errors in their pipetting. Also, the PRO version of the OT-One can hold a single and an 8-channel pipette at the same time. Users can choose a p1000, p200, or p10, and we can consistently transfer 1ul of liquid fusing both single and multi-channel p10.
Actually, most of our users do already have liquid handling robots when they buy an OT-One, and our machine is often labs' first form of automation. Users like this have had lots of success porting their manual protocols onto the OT-One, and you can see the different choices already available for users to download and run on Mix.Bio.
I agree we could use a lot more info on our website, and we are working on it. Till then, happy to talk more details if your interested, just email info@opentrons.com and we can set up a time to talk.
Thanks!
Will
Seems to be a potential solution for small academic/independent labs w/o a budget for what you have labeled a "mainframe" system.
However, you can purchase an old simple tecan/hamilton/beckman for under $2K that has more flexibility then this machine (e.g. ebay "beckman 2000" or "Tecan Genesis 200").
The open source software certainly provides good flexibility for the "hacker" types. But a script step on those other platforms provides the same abilities.
Just giving my thoughts. Curious to hear more.
Thanks for your thoughts, would love to talk more if you want!
Will
0. http://www.hamiltoncompany.com/products/automated-liquid-han...
The OT-One is aiming to be like a 'PC' rather than a 'mainframe' -- affordable, easy to use, and at a scale manageable by individuals. And with Mix.Bio you can download protocols to run, or design them in your browser. It is a lab robot for day-to-day use replacing manual pipetting.
People want information. Lots of it. Pictures and videos. Downloadable PDF's and specifications. They could not care less about these dumb fucking trends that make websites impossible to use.
Sorry. There's just too much of this crap out there. If you understand sales you know this is precisely NOT what your website should be about. Your web developer is costing you sales.
What I mean to say is your comment about a UI that can be improved would have worked without the words 'crap' and 'dumb fucking'.
Profanity is not equivalent to lack of civility. Some of the most well-spoken people have been some of the most profane and brutal in human history. For example, I wonder if Hitler ever cursed in his speeches.
Anyhow, the flowery language was intended to impress upon the owner of this website that there's an urgent need to fix this website because, well, it won't sell very much as it is.
Too many entrepreneurs think selling is done by showing what you make (hardware, software, website). No, it isn't.
I was in that camp twenty-five years ago. I thought I knew how to sell the wonderful hardware I worked so hard to develop. To my credit, I had the good sense to give myself an "F" during a self evaluation of my ability to sell. And that was followed by working hand-in-hand with our first resellers for a full year.
We'd go show the hardware to potential customers and they'd let me do the pitch. After the visit we'd stop at Starbucks for a post-mortem. For the first four to six months they often started with "why the FUCK did you say...".
After a year I got to the point where I could sell our products almost without uttering a word about them during sales visits. I graduated.
Anyhow, from a salesmanship perspective this website is terrible. All it is good for is to mimic current trends in web styling, yet, from a sales perspective it will be horribly ineffective.
When you are trying to sell a product being clever, trendy or sexy seldom works. Having a beautiful woman eat a burger half naked on a pickup truck isn't going to make me want to go to that restaurant. It's a spectacle. It has shock value. I'll probably watch it. But it does nothing to position the product in a useful manner, compel me to learn more and consider buying it.
Building a website loaded with cool and trendy crap is having a website loaded with cool and trendy crap. It is NOT building a website that rocks at salesmanship.
Some of the ugliest ads and websites rock at selling. A/B testing often reveals incredible surprises about how people respond to what they see online.
The OP needs to pay a sales consultant to help them redo the entire thing. It won't be trendy, but it will sell.
That said, the problem isn't that it doesn't work on the iPad. The problem is that you've managed to build a website using what I call "trendy technologies and look" that is, I am sorry to say, absolutely worthless from a salesmanship perspective.
I can't post a fix-it list here. You would be well served paying a sales consultant who knows your industry to help you redo the entire thing. And I do mean the entire thing.
I want to mix up stuff like beard oil or vape juice. Interested in delivering between 0.1ml and 10ml of solution (Eg: 0.1ml of 100mg/ml nicotine, 10ml or vegetable glycerine, or 0.5ml of an essential oil, 10ml of hemp seed oil.)
I never could find a reliable way to deliver small amounts of fluid repeatedly.
In this type of work its the same fluid going to each place so you don't have contamination. I would to produce an arbitrary mix from, say, 40 containers of different fluids.
Would love any suggestions.
I looked at peristaltic pumps and all that, never could find a cost effective solution--- closest I came was syringes attached to linear actuators, and maybe an optical system to observe the position of the syringe to ensure it was the correct distance. OR possibly a squeezing assembly with disposable plastic pipettes, but that seemed imprecise.
I put the idea on the back burner as a result.