COMPUTER MODELING
A model is an encapsulation of knowledge actionable through computer simulation. What would normally take months of lab experimentation to generate insight can be achieved in a day of simulation. Virtual experiments using these models cost much less and enable scientists to hone their choices of the more expensive and time-consuming experiments, thus accelerating R&D.
Prior to computer crash-modeling, about one-hundred prototypes were destroyed to safety-test each new product.
in the automotive industry
The world’s automakers used to perform hundreds of crash tests on prototype automobiles in an effort to improve safety of their designs prior to production. Naturally, building and destroying one automobile after another is expensive, time-consuming and wasteful. A study done some twenty years ago found that a typical physical prototype iteration was taking about 3.8 months to more than 7 months and cost more than US$300,0001.
By 2014, state-of-the-art virtual crash-testing was accurate and detailed.
The same study reported that with computer modeling and simluation, the iteration time dropped to 2.5 days to 6.3 weeks and the cost to less than US$5000, a 50x improvement in both time and cost relative to physical prototyping. No wonder automakers at that time were increasingly making the shift to computer models and simulations to predict how their designs might behave under specific crash scenarios, often running hundreds of virtual experiments before building a prototype and performing a physical test. The speed and efficiency afforded by computer modeling and simulation not just for crash but also for vehicle efficiency and noise reduction transformed the automobile industry. Overall, development time for vehicles has been reduced from 18 months prior to using computational modeling to just 12 weeks. 2
The enormous success of computer modeling in the automobile, aerospace, electronics, and other industries has inspired the creation of the CMMC and its approach for accelerating product development and processes in the cultivated meat field.
applying modeling to cultivated meat
The Cultivated Meat industry promises to mitigate climate change, world hunger, and ocean depletion by displacing fish, poultry, and greenhouse gas emitting livestock with affordable, cultivated meat over the next two decades. Designing products that meet consumer demands and expectations will require copious prototyping prior to mass production.
the problem with laboratory prototyping
Laboratory prototyping is time consuming, and expensive - in both labor and materials. Experiments may take weeks, with mistakes and delays having a cumulative impact on limited budgets when striving to be the first to market.
Some of the factors that cultivated meat startups need to harmonise to create the “holy grail” of cultivated meat that consumers will want to buy - a steak.
The solution is to use Computer Aided Design (CAD) to hone lab experiments using faster, cheaper, virtual experiments. Just as in safety crash testing in the automotive industry, computers are great at managing complexity and creating models that encapsulate and categorize knowledge in the life sciences. This allows biologists to reach insights faster and more efficiently. If we take the growth of skin tissue as an example, Biocellion SPC, one of CMMC’s members, reports that customer experiments taking 3 weeks in the laboratory and costing hundreds of dollars can be simulated in less than a day and $25 of leased computer time.
a range of possibilities
Simulation of young, healthy skin growing on a scaffold in vitro.
The base knowledge and technology for applying computer modeling to the science behind cultivated meat already exists. Models that deepen understanding of how new skin tissue grows, its elastic properties, and its barrier function, have all been built with virtual experiments. These virtual models have been validated against physical experiments.
Going back to cultivating a steak: imagine being able to modify nutrient concentrations, scaffold designs, and temperature, while observing the impact on growth rate, the patterning of different cell types (marbling), elasticity (mouthfeel), and water content (cooking qualities), in just hours of computing time and without waste. Imagine computers exploring the enormous space of growth process possibilities, providing just the optimal configurations for final verification in the laboratory. Imagine getting the affordable high-quality cultivated meat products consumers demand to market years sooner than would be possible without computational modeling — that is our vision.
1Stark, A. (2019, February 15). How Simulations Drastically Reduce the Development Time of Cars. Retrieved September 28, 2019, from SPOTLIGHTMETAL website: https://www.spotlightmetal.com/how-simulations-drastically-reduce-the-development-time-of-cars-a-797868/
2Spethmann, P., Thomke, S. H., & Herstatt, C. (2006). The impact of crash simulation on productivity and problem-solving in automotive R&D (No. 43). Working Paper.