Open Hardware Project: Interview with Vadim Kimlaychuk

8 min read

Open Insulin is working to make an open source protein purification system (FPLC), which is a key piece of equipment for manufacturing insulin, and one that is currently very expensive. Avery McLain spoke with Vadim Kimlaychuk, one of the lead engineers working on the FPLC project, about his work, as the first installment in a series of interviews we have planned with project participants and stakeholders.

Please introduce yourself and tell us a bit about your background.

My name is Vadim Kimlaychuk, I’m 43 years old and I live in Estonia. My current occupation is IT infrastructure manager, working for the big provider of the card payments called NETS. My background is in system engineering. I graduated from the Moscow Institute of Electronic Engineering with two master’s degrees, where I specialized in Information Security and Computer Systems and Networking. I also have a PhD in System Automation from Tallinn University of Technology in Estonia.

I’ve lived all my life in Estonia, but with Russia we are close neighbors and there were at least a lot of connections before. Now it’s a bit tough to keep them. My interest in this project is connected to my state of being – I have type 1 diabetes and it was always interesting for me, what can I do for this state of being to make my life easier and to make the lives of other people with the same problem easier. That’s how I came to the Open Insulin project. Maybe a little bit later than expected, because I’ve had this problem for nearly 20 years. Since I have a technical background, I started to participate in our FPLC project that is connected to protein purification system design. At the moment, I’m nearly 2 years with Open Insulin.

What is it like working remotely on a global open source project?

It didn’t go well all the time, it’s like waves you know, you have something that is pushing you for a short while and then a period of silence for gathering information, and then you have a step forward. Sometimes you even step back to revise and change the previous work due to some critical components missing on the market or requirements have been changed. We depend on specialized hardware a lot. Since there is no strict plan what to do and when and resources are limited the overall progress is very slow. So, for the open source project I participated in, I would say more connected to software development. There are some projects on github, java programs and other collaborations, but they were not connected to Open Insulin particularly. Hackuarium is another OSS community I helped to develop software for the “corona detective” project last year. Developing software is easy nowadays. Have more freedom, easy to collaborate and track changes. Clear goals and technologies.

The FPLC project is one of the projects I worked on as an open source project in my spare time, because the rest of the time belongs to my family and my work that I also have.

What is the overall purpose and design of the protein purification system that you’ve been working on?

The overall system design is taken from open sources, because the general idea is simple in fact. We have several parts like pumps and sensors connected together. It’s not a secret, it’s not any kind of intellectual property that we need to be aware of, and we are basically implementing the general FPLC design everyone can find on Wiki.

FPLC stands for Fast Protein Liquid Chromatography. We are going to use it for insulin protein extraction in the final phase of the production. Professional systems on the market are very expensive. Cost more than 100k and are not affordable for the enthusiasts. Our goal is to build a similar system from the components that are accessible on the market and keep the net cost less than 2000 USD.

Yann, our biologist, has practical experience with such systems that is why he defines requirements for the project and describes components. Overall design is his creation.

What are some of the challenges of everyone working on a different part of this but none of you are physically together?

Yes, one of the challenges is that those parts are not available in ordinary shops. You need to find it and order it from anywhere on the Earth. We spend a lot of time finding exactly what we need. The conductivity sensor, for example, we took from a water purification system. It’s not designed for FPLC and we are adapting it to use it, but it’s not exactly what we need. Conductivity sensors for FPLC may cost starting from 2000 USD and this is the only sensor without a control circuit, without calibration and without a software driver. Very expensive.

That’s why for example, this conductivity sensor is also designed by ourselves. Zach is thinking about a new sensor that he might produce by himself, and he will test it and then calibrate it and so on. We try to build this system as simple as possible from parts that anyone in the world can get. That’s tough to follow.  

What is some of the current progress you’ve been making and what are you planning to work on next?

We continue to mark time. There are active people in the projects and we have free will to choose the part or module to work on, but the lack of coordination and project management leads to some kind of working loops. In general we are still on modular tests. We have agreed upon the process of module stabilization lately. The FPLC project has a fixed number of modules and when a module is assembled and tested by 3 developers it gets the status of “fixed” for final assembly.

Mechanics are very important, I think. Mechanics and overall mechanical and functional design must conform to each other. When all modules are ready for the final assembly we need to put everything together, and that’s hard. We didn’t think much about the final assembly. For me, it seems like simple stuff and I tried to assemble modules together at home and I failed. I completely failed there, and that’s a bit depressing for me because I think that the final design and the final steps to put everything together will be one of the simplest steps, but unfortunately not.

So if people were interested in getting involved with this or wanted to help out, are there any skills or anything in particular you’re looking for help with?

I think a mechanical engineer, at the moment, is the one we are missing. Zach has some experience with developing pumps, he has provided us with the syringe pump design, but it’s hard to reproduce for me. We also have a peristaltic pump designed by Steve. The same problem. So yes, the mechanical design is a blocker at the moment.

We need more proof from the other community members that they have succeeded in building modules separately. Now we rely only on a single person that says “okay, this module works for me”, but to be able to finalize the project, we need some kind of assurance that it will work for everyone, not just for a single developer. As I wrote before for me, it works with flow cell, peristaltic pump, conductivity sensor, pressure sensor,  works for me, then we are arguing about mixing chamber that is not ready, and again, it’s basically standing behind mechanical design because Ian said “yes yes yes it’s very simple, we just need two more magnets and the motor and that’s it”  Everyone understands that it’s supposed to be simple, but you need to build it, you need to connect it to the controller, and to write at least a test program because every sensor, everyone component is followed by a small piece of a test code that is put into an arduino circuit.

The Raspberry Pi plays the central controller role. It is our final assembly part, when we connect all the modules we approved and tested to it. We assemble everything together and the final design will include writing and testing the program for the Raspberry Pi I would say. Because by that time, all the modules should be ready and physically the system will also be ready, but before that, we need to approve within the community each and every module works by a certain number of developers.

The final design of piping, tubing, and components layered within some box, there again it looks quite simple. If you take a computer, it’s easy to understand how it works, you have a monitor, a keyboard, a mouse and so on. But if you start to think about details, about how all is put together, what are the measurements, what is the material it’s made from, then you need certain skills to design it.