For centuries, people have been striving to make sense of the nature in which they live. Classical physics was explained using linear and deterministic formulas up until recent times. Isaac Newton, who is considered the father of classical physics, invented calculus to model natural and linear systems. At that time, these linear systems could be measured and tested by slightly adjusting the input states. As a matter of fact, much of the scientific knowledge we have inherited was obtained through these tests — in fields like chemistry, which is used to produce medicines, or biology, which is used to explore nature or diagnose common diseases, along with other scientific knowledge we have been using for centuries.
However, it has become apparent that this situation has changed since the advent of quantum physics. Some systems have become increasingly sensitive to their input states, making them difficult to test through simple input adjustments. The old formulas used to solve linear and low-entropy problems in classical physics are now falling short of explaining such complex and high-entropy systems, which are also called post-Newton paradigms.
After the advent of quantum physics in the late 20th century, working with nonlinear and non-deterministic systems brought challenges in testing complex systems. This new world of quantum physics is seeking new solutions to reduce the entropy of these new complex systems. Like other technological innovations that make our lives easier, this problem will lead us to further innovation.
Thanks to computer technologies pioneered by Alan Turing, who famously cracked the Enigma code during World War II, we have introduced a new approach that not only builds on the methods used to solve linear and deterministic problems since Newton's era but also enables the solution of more complex and chaotic problems.
Parallel to the advancements in quantum physics, computational technologies have significantly evolved. Advancements in transistor technologies, as predicted by Moore’s law, have led to the development of modern CPUs. With the high computational power of these new CPUs, people have been able to model and test complex problems of the modern world, such as weather forecasting, economic management, environmental effects of climate change, and many other complex issues.
Classification of Simulation Models
Simulation models used in modern research can generally be classified into two main categories: deterministic and stochastic.
Deterministic Models
Deterministic models produce the same output for a given set of inputs. They do not involve randomness, and the outcome can be predicted in advance. These models can be divided into:
-
Static deterministic models
Used when input values remain constant over time.
Example: calculating the load-bearing capacity of a beam under a fixed weight. -
Dynamic deterministic models
Used when input values change over time, either continuously (e.g., fluid dynamics) or discretely (e.g., production line capacity).
Stochastic Models
Problems that involve randomness can be modeled with stochastic models. These involve uncertainties and can produce different outputs using the same inputs.
They include:
-
Static stochastic models (Monte Carlo simulations)
Used when time is not a variable.
Example: risk analysis or portfolio management. -
Dynamic stochastic models
Time is a variable.
Example: modeling emergency service line status or stock price movements.
This classification system helps researchers and engineers select the most suitable type of simulation based on their research requirements.
Simulations in Real-World Applications
During the COVID-19 pandemic—which caused over 100 million infections and 2.4 million deaths—researchers were urgently developing vaccines as every day was critical for global public health. Developing a vaccine was a challenge, but distributing it worldwide was also a significant challenge. Vaccines had to be stored at extremely low temperatures and transported through reliable cold-chain logistics networks.
Before simulation technologies, researchers were unable to develop solutions this quickly for such a complex problem. Thanks to simulations, they were able to create virtual environments to model this challenge and accelerate development in a safer and more agile way.
Simulation plays a critical role in defining, designing, optimizing, and validating future real-world products in the virtual world. Through simulations, we can significantly shorten the design cycle and create lighter and more reliable products.
With recent advancements in computational resources—specifically modern GPUs with high core counts and parallel processing capabilities—researchers and engineers can build more detailed virtual environments and model high-entropy, complex problems efficiently.
Recent developments show that NVIDIA, the world’s largest GPU manufacturer, is planning to use a simulation technology called Omniverse to train its real-world robotic products. By implementing Omniverse into their production pipeline, NVIDIA aims to reduce training costs and overcome hardware limitations through virtual environments.
The Future of Simulation Technologies
The future of simulation technologies is playing a critical role in modern scientific research. With further advancement in technologies such as quantum computing, researchers and engineers will access cutting-edge computational resources that traditional computers have not been able to provide throughout history.
These new capabilities will allow modeling some of the most complex problems, such as calculating anomalies in climate change and understanding the mechanisms behind them. Hopefully, with these improvements in modern science and the enhanced ability to model complex systems, we will be led toward a greater future.
References
- Carson, J. S. (2005). Introduction to modeling and simulation. Proceedings of the Winter Simulation Conference, 2005, IEEE. https://doi.org/10.1109/WSC.2005.1574524
- Sun, X., Andoh, E. A., & Yu, H. (2021). A simulation-based analysis for effective distribution of COVID-19 vaccines: A case study in Norway. Transportation Research Interdisciplinary Perspectives, 11, 100453. https://doi.org/10.1016/j.trip.2021.100453
- Winsberg, E. (2013). Computer simulations in science. In E. N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Fall 2013 Edition).
- Aybars, A. (2021). Karmaşıklık Ekonomisi: Post-Newtonyan Paradigma. Scala Yayıncılık.