If we want to assess if a phenomenon is going in a direction we expect, we can simulate it, using digital models, possibly agent models factoring in known properties and interaction features.
We can alternatively, " go live" in the real world and observe what happens in realistic, concrete, but still bounded and measurable conditions.
How do we decide between one or the other?
-skill availability for sketching a model and programming a simulation environment and digital experiment?
-risk/cost/liability to be met in real-world experiments?
-science skills? Field experiment aspects, digital modelling aspects?
In my experience on the digital antenna arrays, the main factor is financial recourse for experimentation. Digital twin technology is more cost-effective and easier. In this regard the more than 10 last years I use digital experiments for researches in the area of digital signal processing, antenna synthesis, etc. But by the synthesis of new metamaterials, only the experiments can give a possibility to confirm the negative refraction index, because the current versions of electromagnetics simulators should be improved.
Simulations merely require the existence of an abstract correspondence between the simulating and the simulated system. In experiments, in contrast, the causal relations governing the experimental and the target systems are grounded in the same material.
The experiment is valuable because it gives a physical connection. The simulation prepares for the experiment. In the Covid situation, experiments are more difficult, downright impossible.
It depends on the cost of each of them and on the efficiency of each of them in giving results that are close to reality
Dear Di Francesco, thank you for sharing very interesting question.
I agree with Dr Vadym Slyusar..
Kind regards...
Many thanks for the interesting views especially the multifaceted approach of @Vadym Slyusar, @Teodora Hristova and @Faraed Salman.
More details and comments welcome...
Experiment is better, but may be demanding and requires time and technicalities. Simulation predicts the expected outcome in the models.
Simulation is rarely in the position to replace the measurements. It has, however, one big advantage: we can freely change the parameters in the model to see which are more important than others, which influence more strongly the result. Another advantage is that in macroscopic measurements we see the average, while in, e.g., molecular dynamics calculations we can see the microscopic details. A third advantage is that simulation makes it possible to study the properties of materials and structures that cannot be easily synthesized, or which require extreme conditions or special technologies. Possibly the simulation is not exact and complete, still, it allows reducing the number of experiments and to select the most promising solutions in advance. So, I think, the proper place and use of simulation should be found: not replace experiments, but to complement them. To understand better phenomena and to reduce the number of expensive experiments.
Dear Dr. Di Francesco!
You raised an essential point to consider. I see simulation as a part of experiments:
1) Korani, W., Vaughn, J.N. Crossword: A data-driven simulation language for the design of genetic-mapping experiments and breeding strategies. Sci Rep 9, 4386 (2019). https://doi.org/10.1038/s41598-018-38348-y Open access:
https://www.nature.com/articles/s41598-018-38348-y
Simulations can give the parameters needed for designing a specific experiment later on:
A case-study: Yaroslav Vyklyuk et al. (2021). Modeling and analysis of different scenarios for the spread of COVID-19 by using the modified multi-agent systems – Evidence from the selected countries, Results in Physics Volume 20, January 2021, Open access:Article Modeling and analysis of different scenarios for the spread ...
Yours sincerely, Bulcsu Szekely
The experiment is at the head of the simulation. You can simulate on paper, but only experiment can confirm
Simulation in the lab first to test efficacy followed by a real experiment in actual environment to test effectiveness.
Dear Renaud Di Francesco we work in the field of synthetic inorganic chemistry. Thus our research work takes place exclusively in the laboratory. Personally I cannot imagine how this could be done by simulation. However, the corona pandemic has triggered new ideas even in our discipline. Please see this interesting link entitled
Virtual Chemistry and Simulations
https://www.acs.org/content/acs/en/education/students/highschool/chemistryclubs/activities/simulations.html
Softwares are currently being designed to handle more complicated methods. The way I see it, in a couple of years down the line, 95% if not all fields may have to simulate, then make some conclusions before evaluating the possibility of carrying out actual experiments.
It is better to simulate and compare results with experiments.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators