Many labs across the world are racing towards two goals right now: 1) developing COVID19 vaccines, and 2) developing immunoassays (test kits that verify the immunity of the patient). The next challenge, however, will be to produce them on a scale enormous enough to supply the entire population. For most parts, the production will take place in bioreactors - starting from small lab flasks, going all the way to industry-grade cistern-size reactors. Although the technology has been around for decades, the process of scaling from smaller reactors to larger ones is highly complex and non-linear, therefore, being time and resource demanding. Finding the optimal bioreactor growth factors can take up to 12 month and costs millions of dollars, thus prolonging the time-to-patient of vaccines and immunoassays and making them less affordable.
What it does
When vaccine development phase is done, they get mass produced with the help of bioreactors. We are building an optimization tool for these bioreactors(advanced design of experiment software) to accelerate the COVID19 vaccine and immunity test-kit reagent go-to-market time. The software helps in finding the right bioreactor process and microorganism nutrient mixture parameters for scaling the production - rapidly and with less resources required.
How we built it
We are building a bioreactor process optimization tool named "xT BioE". By combining the industry standard Design of Experiment (DoE) methodology, that has already been proved for many years, with innovative AI algorithms and our optimization expertise in the fields of additive manufacturing and specialty chemicals, the team has embarked on a mission to deliver a state-of-the-art software for finding the right bioreactor process parameters and cell culture nutrient mixture parameters. The software will help to scale the production of vaccines and reagents, shortening the time needed to move from a smaller bioreactor to a larger one. The SW uses active reinforcement learning and small data AI approach with an iterative methodology for finding the optimal machine and mixture parameters rapidly and with fewer tries. We are building the graphical user interface with an ease-of-use in mind in order to make it as simple as possible and decrease the on-boarding time, so that no second is lost. The solution will directly benefit COVID19 patients and general public, as it will make the infection rate more manageable sooner. We intend to provide both the software and our human resources for FREE OF CHARGE to any lab, biotech startup, production, scaling and biopharma company in the world to use it for COVID19 related development. We already have a working prototype ready, the MVP should be done by the end of April, and by the end of May first pilot customers will be using it.
Challenges we ran into
- Accessing research data sets to test our algorithm on
- Very few people actually know what we are talking about - very obscure and industry specific problem, yet a topical one
- Hard to validate our business and market assumptions in a quarantined world
- No vaccine production labs in the local vicinity
- Hard to access key people in the industry
- Challenging to communicate with people involved in the vaccine development now because of the emergency situation ## Accomplishments that we're proud of
- Expanded our team with very highly skilled and rare competences
- Got some recognition and collaborations
- Are building a project that should contribute to getting out of the global lockdown sooner ## What we learned
- To harness the full power of mentoring sessions
- To work on communicating our idea with a lesser degree of complexity so that a wider audience understands the gist
- The financial resources are there, just have to work a bit harder to get them
- Use the situation to attract the talented people sitting at home ## What's next for xT Bioreactors 1) Running lab tests with the Institute of Microbiology of University of Latvia 2) Delivering the MVP 3) Piloting with customers 4) Rolling out the full functionality product