What if I told you, we can predict the future with > 90% accuracy!
This is no less than using a time machine to visit the future and foresee outcomes.
While we are still away from inventing Time Machines, we are close to harnessing the entire potential of another technology, which can help us predict the future.
This another technology is ‘Digital Twins’
Gartner’s emerging technology radar 2023 identified Digital Twins to be a “Very High” impact emerging technologies in the coming “1 to 3 years”.
What are digital twins though? And what is it capable of?
Let’s deep dive…
Introduction:
During the Apollo 13 mission of 1970 by NASA, no one could predict that the oxygen cylinders would explode early in the mission, bringing survivability an outcome of few decisions to be made 200,000 miles away on earth.
Fortunately, NASA had the Apollo 13’s “Digital Twin” model, which was used to look in the future as it simulated different courses of action that could be done to resolve the matter with the least amount of after-effects.
Though Apollo 13 did not land on the moon, it was successful in bringing back astronauts back on earth.
As the name suggests, digital twin is a ‘Digital’ replica of a real world object. This object behaves similarly to the real object in the virtual world.
What if global warming heats the earth by more than 4 degrees?
We know the outcomes would be catastrophic. However, what? Then why? What else can we do in those circumstances?
It is almost impossible to precisely evaluate the problem without heating Earth by 4 degrees. But, it is possible to create a digital representation of Earth and test out our theories there in the virtual world in order to understand the results in the real world. (This also being a reminder of the increasing global warming, and the steps we MUST undertake to prevent that from happening).
How it Works:
A trained virtual model that mimics an object’s real-world behavior in the virtual world is known as a “digital twin”.
Stunt doubles, who dress identically to actual heroes to execute stunts on their behalf in movies while safeguarding them from the risks involved, is a good use case, that throws light on digital twins.
Digital twins are created to take on the characteristics of real-world items, just the way stunt doubles adopt the characteristics of real-world heroes. Combining three data fundamentals allows for this:
- Situation data
- Object data
- Physics
While object data helps in defining the digital replica, situation data assists in building the environment that will host it. Finally, both are synchronized with the laws of physics to derive analysis.
To better comprehend this, consider putting a Honda Civic to the test on the terrains of Ladakh.
The information needed to build a copy of a Honda Civic comprises the weight, aerodynamics, wheel size, engine power, and other details.
Whereas Situation data, as in — temperature, altitude, road conditions, oxygen levels, and other factors, are required to generate the Ladakh environment.
Both the above are synchronized using physics. An example — higher steepness of the road > higher engine power to climb > higher engine wear > decreased engine life > decreased car life. One of the zillion possible outcomes.
Many simulations, like the one just stated, are modeled using algorithms to produce precise data. Additionally, the generated data is examined for correctness against real-world data obtained by IoT sensors.
Once the model is validated, voila! We can generate insights into anything and everything. What if it snows that day in Ladakh? Would my Honda Civic be able to make it?
Types of digital twins:
Any tangible or intangible thing from the actual world can be modeled in the digital world. As a result, rather than being categorized according to the target item, digital twins are instead categorized according to the use case of the digital worlds model.
We broadly classify digital twins into the following four categories:
- Component/Parts Twins: This comprises examining specific components of a larger asset. Consider how a car’s clutch pad performs under various driving scenarios.
- Asset Twins: The asset is made up of parts and components. Understanding how a car’s various components, such as the clutch pads, brakes, driving style, engine, and others function as a whole makes the basis for understanding assets.
- System/ Unit Twins: Assets form the building blocks of a system. What if we wish to analyze the traffic system on highways to better manage them? This calls for collective analyzing of the individual assets like cars & driving behaviors to foresee breakdowns and inappropriate driving, leading to traffic.
- Process Twins: Systems enable processes. What if a highway needs to be closed for a few days? This calls for understanding the backup highway systems for if they can hold the diverged traffic amongst other possible outcomes of the action.
Digital twin is an emerging mega trend, and to amend this mega trend into our lives calls for understanding what it is and what it is capable of.
During this note we deep delved into:
- History of Digital Twins
- What is it?
- How does it work? &
- What are its different types?
Coming along in the next part of the series we’ll further delve deeper into its use cases, potential, approach and other aspects we must know…
Author:
Kanuj Jadwani
