Meeting summary: Tech x FAW group discussion 3: automation for cultivated meat

[chadbrouze]

Thank you to everyone who joined yesterday's discussion. Here are some notes. Feel free to correct me, add to the list or share lingering thoughts in the comments.

Takeaways

• Automation's primary value proposition may be risk reduction (contamination, operator safety) rather than labour savings
• Inter-machine incompatibility suggests value in orchestration not just task automation.
• Technician pain points (e.g., gowning time) and safety concerns (chemical burns) indicate adoption may align with operators.
• Cultivated meat challenges strongly overlap with pharma-scale bioprocessing.
• Factory-style automation makes sense pilot->production scales since processes are fixed and more flexible robotic solutions make sense R&D->pre-pilot scale

Promising Domains

• Sample taking can be time-intensive and introduces contamination risk
• Post-processing (after batch harvest) makes operators spend 10-12 hour days in cleanroom. It includes many process steps with strict regulatory oversight.
• Bioreactor Cleaning
  • Dishwasher style machine exist for small parts at small scales where bioreactors are disassembled after each run.
  • Larger systems usually use CIP (dangerous chemicals) or SIP (pressure tolerance requirements). Difficult to use UV due to crevices. Although bioreactors are now being designed around UV cleaning
• Chemical / Media Preparation
  • Companies receive high concentrate products from suppliers and then dilute in-house. This is a safety risk (chemical burns) as well as a contamination risk.
• Sensor and control systems are essentially trying to solve homeostasis outside the body. Can we borrow insights from other scientific domains?

Tele-Operating Robots

• This technology is currently transitioning from research to industry.
• Promising at small scales where processes are not set.
• Potential double win. Reduced contamination risk while collecting training episodes for imitation learning.
• Many companies automate individual tasks but humans must move items between stations. Also sometimes bottlenecked between lack for compatibility between machines. For example, one machine wanted a 96 well plate and another wanting 48 wells. Could an arm on wheels solve this?
• Stopping operators from coming into a manufacturing environment may be embraced by technicians because gowning in is a pain (~10 minutes in and ~10 minutes out).
• Could reduce contamination risk for media preparation.
• May improve safety for any process that involves harsh chemicals.

Who Could Pay for This

• Pharma faces many of the same problems but the problems may not be prioritized. For example, pharma doesn't care about media costs because dwarfed by regulatory costs
• Technology could be developed at cloud labs? What is the bottleneck for increased adoption of these facilities.