The Difference between Biomimetic Digital Twin Ecosystems and Mechanistic Digital Twins
Courtesy of Joe Glick, Chief Innovation Officer – RYLTI
“In the newspeak of IT marketing “Digital Twin” has joined “AI” in the buzzword twilight zone of business leaders. A recent Capital One study “Discovering Data Management” reported business leaders found it difficult to understand their data.
These concepts and phrases have been around for years, and the technology has proved an effective solution for many challenges related to monitoring and maintaining engineering products. But now it has become a buzzword in the domains of life science and healthcare, and that is a concern.
Mechanistic digital twins (DT) can help to ensure that medical treatment technology is performing correctly, but the more important questions are
- Is it the right treatment for this condition?
- Is it the right treatment for this patient?
The answers to those questions are related to a hard reality: what we don’t know about the complex relationships in biology is mind-boggling. What we do know is that this complexity is created by ecosystems of various scales and domains, which interact internally and externally, multiscale and multi-dimensionally, and if that wasn’t enough they adapt dynamically. The conclusions:
1-Discovery and learning supported by real-world evidence are the most important objectives to advance precision medicine. In biopharma, the vast majority of time and money is spent on paths that eventually prove unsuccessful. Imagine discovering early evidence of fruitless paths!
2-The discovery methodology must be biomimetic, imitating real-world ecosystems by modeling the interactions between agents and independent components of the modeled problem domain.
3-Using mechanistic digital twins will not enable discovery and learning. The differences between the two technologies are summarized in the image.
Digital Twin (DT) Technologies
Biomimetic DT Ecosystems
- Model agents or independent components of a problem domain that may interact
- Output real-world evidence of previously unseen interactions
- Biomimetic design supports interactions between models of diverse domains and scales enabling discovery & learning
Mechanistic DT Methods
- Model defined systems or subsystems
- Simulate or predict the behavior of the model in diverse scenarios
- Mechanistics design limits the learning from interactions between models of diverse domains and scales enabling discovery & learning