Hey, i submitted my application for the posted PhD program but wanted to follow here as well. I hoped to start a conversation about amybo as I just found some videos posted 2 years ago regarding the pioreactor. I think this open source modular control system I’m working on would help amybo moved towards a sustainable open source future and I would love to chat more about how I could help. If there’s any Information you could provide me regarding how someone with experience in biology, 3D printing, computer science, and making things could help your efforts, I would love to learn more.
1 Like
Great to hear you applied for the PhD. And yes, a conversation on how you could help with AMYBO would also be extremely welcome.
Since I don’t know what your modular control system does, I’ll outline our current work so you can determine if it fills any of our gaps. I should say that none of this is currently intended to be part of the scope of the PhD program, so I wouldn’t focus on it too much in your interview.
We currently have two electroPioreactor models: the original Mixed-culture electroPioreactor (MEP), a lower-cost version currently being trialled by an MSc project at Edinburgh University, and the Aseptic electroPioreactor (AEP), which Imperial and Edinburgh PhD students are working with under the CARMA Hub project (which this PhD is intended to partially continue).
The Pioreactor software (Open Source, developed by @CamDavidsonPilon) is very good at controlling bioreactors. Our electroPioreactor plugin adds electrolysis and CO2 sparging control.
In addition to the standard Pioreactor, and controlling other bioreactors (as done by companies like https://changebio.uk/), the HAT has just enough ports to meet our current minimum requirements: pumping media in and out (in chemostatic mode), controlling stirring, and controlling CO2 sparging uses its four PWM channels. It has four LED channels one of which provides IR light for DO control (in turbidostatic modes) and another controls electrolysis through a spare LED output, two LED ports are free in our current setup. It also controls temperature through a separate PCB and has a StemmaQT / Qwiic connector which we don’t yet use in our official builds, but I connected an Adafruit spectrometer to mine.
I’m mentioning @gerrit specifically, as he may well be interested in your open source modular control system and can likely correct me if I’ve missed any Pioreactor capability.
For future versions of the AEP we are potentially looking at further I/O which may go beyond the current Pioreactor HAT’s capabilities:
- Potentiostatic control of a three electrode system - possibly using a Rodeostat.
- pH measurement and control
- O2 and/or CO2 measurement and control
- H2 interlock to prevent electrolysis unless the detected H2 concentration is below the LEL
- alt Media pumping - while this is a standard Pioreactor feature, there are only 4 PWM chanels and our CO2 relay currently requires the alt Media PWM
Wishing you all the best with the PhD application, and I hope we’re able to work together however it goes.
Thank you for all this information. Considering the Pioreactor project is opensource, i am wondering if i may be able to adapt it to serve as a front end for my system. I have tried to build a front end but quite frankly im much more interested in the back end and microcontroller code so i havent tried very hard. i ask claude to try and make something and it returned nothing of use. Regarding my system, i have been working on it for quite a while and i am on version 3 right now. The intention is to provide a cheap and powerful system based currently on the esp32-s3 that can be adapted to a specific use-case with modules. The base system is comprised of an interface layer (serial or mqtt) command system / handler, generic controller daemon, calibration system, and among other misc things a worker system that allows users to configure multiple MCU’s to communicate and be controlled by one host via espnow, i have designed the worker system to be a smaller build of the main system so code is shared heavily. The intention of workers was for IO expansion, and the ability to add multiple sensors and effectors around a location such as for gardening, outdoor monitoring, temperature mapping, etc. A module can be anything from a sensor, an effector (motor, solenoid, smart plug, etc), other misc code such as a way to send data to a database. I have yet to fully implement and test the controller system, it currently has basic control functionality but not the full modular sensor and effector relationships i hope to reach. Currently im creating a pocket sized bioreactor to bring along to interviews in hopes increasing my chances of getting a new job, i also have a refrigerated incubator built from a mini fridge thats been working pretty well (need a solution for temp fluctuations to bring them down from ± 1C), i put together my full bioreactor PCB and have yet to design a new mason jar bioreactor for that, i’m putting together an automated plant pot, and also automating my garden. These various use cases have been quite easy to test and build because the system is designed as modular as possible so i can just enable and disable features as needed. Regarding how it could be of use to your work, once i have the controller logic stable and functional i believe if it can be measured and effected, the system can manage it. I would like to report back when the control and calibration system is working as intended on how it could be of use. Regarding your alt media pumping, i have built and tested this system as i intend for my bioreactor to use it to automatically determine optimal media recipes: Cam Valve - Gen3 . Further more, regarding your pH conundrum, i am having trouble sourcing a cost effective autoclavable pH probe, so i intend to try and implement a spectrophotometric system as i want to build a microbioreactor eventually to automate research, please see this article: Redirecting. Sorry for the lengthy response
Thanks for such a detailed write-up - there’s a lot here that overlaps with where we’re heading. Using a Pioreactor as the front end to your ESP32-S3 system is interesting, I understand the Pioreactor UI/MQTT layer is open, so in principle it could drive or display your worker nodes. Worth a proper look once your controller logic is stable - do report back when it is.
I personally find it easier to comprehend a project when it is expressed in terms of real-world benefits rather than abstract means of achieving those benefits. What do your worker nodes do in the context of an AMYBO electroPioreactor experiment?
I can definitely think of uses for your mini-fridge as we took a huge one up to do loch sediment sampling just a few weeks ago. One that could take say 15 50ml skirted centrifuge tubes and maintain them at 4°C for 12h in an EV without requiring range-depleting power, could be really quite useful.
The Cam-valve-gen3 link is timely: we had a Harvey Mudd team working on a pinch valve, they didn’t quite nail it, so @Gerrit rejuvenated the project with this competition: FreeCAD Community Design Contest 2026: Open Science - FreeCAD News
And the spectrophotometric pH idea is also interesting. Personally, I’d use other ways of sterilising the pH probe than autoclaving, but the real issue we face is getting a cost-effective probe that we can fit in the tiny Pioreactor vial. The problem with spectrometry for us would be that you need to dose in the indicator solution, which would likely have adverse impacts. If you do it instead on a sidestream, or the outlet, you could just use a regular pH probe, if the flow was large enough. However, the system you propose could be used for any colorimetric test. IF you could do it with a small enough sample volume, then it may actually be very useful for pH monitoring on the outlet.
Of course, there is much to say so i did not organize it particularly well. The worker nodes i don’t see having much use in the context of the electroPioreactor, they are more of an add-on to make the system more useful in situations where its not feasible to add communication infrastructure to all the locations that need be measured or controlled. I didn’t have too much direction when i added them except that it might useful for field research with a battery and / or solar panel.
Regarding the mini-fridge, I am not too familiar with the specifics of the refrigeration cycle which i believe would be needed to achieve that level of efficiency. the issue i’m running into with my build is that the compressor has a lockout period after its been shut off that leads to significant heating from the 100W of lighting in the fridge. If we were able to reduce this lock-out period then achieving tighter control would be much easier. This of course is only a problem for a photoincubator, for your use case a very well insulated vessel might be suitable with some ice packs. If you’re leaning more towards the fridge idea, i’m not sure on the power draw of a mini fridge but they may already be efficient enough for this use, especially if you get past the losses from converting to AC somehow (dont convert, DC fridge?).
Regarding spectrophotometry. i want to make a compact small volume flow-through spectrometer for this use case and i have some ideas but it is not a priority for me yet unfortunately. I would use the fluid-valve system to mix a sample of media with the indicator solution and then measure before purging the line. i came across these pH sensor spots sometime ago but the only option is to request a quote, so i assumed it was far too expensive: https://www.presens.de/oemcustom/detail/ph-sensor-spots-sp-hp5 If we are able to replicate this technology that might be ideal, atleast for my dreams of a microbioreactor (<1mL wells).
These are all dreams i have that will eventually come to fruition, once i am happy with the main system. i’m quite busy recently but once i have some more free time i’m very interested in replicating your electropioreactor setup in my system and seeing how i can contribute to your efforts.
Thanks again, very much looking forward to seeing your electroPioreactor setup and how you improve on it.