All articles

Digital Twins

Every year, the sphere of technology represents new functions or gadgets that not only simplify life, but also increase the turnover of goods in the market at times. According to the forecasts of research company Markets and Markets, by 2020 the profit from sales of the artificial intelligence franchise will grow up to 5.05 billion US dollars. For comparison, it should be noted that in 2014 the profit was 420 million dollars.

Competing companies Apple, Microsoft, Samsung and Amazon released voice assistants, in demand by 40% of people around the world, and the founder of the corporation Acceleration Studies Foundation – John Smart, launched a project called “Digital Twin”, which should be smarter than its predecessors.

The idea is that copies of things used in everyday life will be created, which will be able to anticipate, predict and analyze the basic properties of the human personality.

Who are the “digital twins”?

It is a mistake to believe that the term “digital twins” refers to robots and artificial intelligence in the guise of some kind of humanoid being. The term itself is currently used for the most part to industrial production. For the first time the concept of “digital twins” appeared in 2003. The term came into use after the publication of the article “Digital twins: excellence in production based on the virtual prototype of the plant” by professor and assistant director of the Life Cycle Management and Innovation Center at the Florida Institute of Technology Michael Greaves. The very concept was invented by the engineer NASA, who was a colleague of the professor.

In essence, “digital twins” is a concept that combines artificial intelligence, computer training and software with special data to create live digital models. These “digital twins” are constantly updated after the change in physical prototypes.

Where do “digital twins” get  the data for self-renewal?

A digital copy, as befits artificial intelligence, constantly learns and improves. To this end, the “digital twin” uses knowledge from people, other similar machines, from larger systems and the environment of which it is a part.

Michael Greaves has proposed three of his requirements, which must correspond to the “digital twins”. The first is the correspondence to the appearance of the original object. It is necessary to understand that a similar appearance is not only an entire picture, but also the correspondence of individual parts to a real “twin”. The second requirement is related to the behavior of the double during the tests. The last and most difficult is information that is obtained from artificial intelligence about the merits and demerits of a real product.

As Michael Greaves notes, when digital copies were put into use, even the criterion of external similarity was considered difficult to implement. Today, once the digital double is identical to the first parameters, it can already be used to solve practical problems.

“Digital twin” will be able to fully learn about your habits, features and hobbies, and with the help of programmed intelligence to search for people with identical desires, which, perhaps, simplify our life, making it more exciting. For this, the “digital twin” will use an algorithm that collects our data from social networks, cloud services, personal correspondence and the whole Internet. Therefore, it is worth considering how much we are willing to entrust such data to the “Digital Twin”.

Why do we need “digital twins”?

Digital copies are created in order to optimize the operation of physical prototypes, entire systems and production processes.

According to Colin J. Parris, Ph.D., vice president of software research at GE Global Research Center, “digital twins” is a hybrid model (both physical and digital) that is created specifically for specific business purposes, for example, predict failures, reduce maintenance costs, prevent unplanned outages.

Colin J. Parris says that when we talk about “digital twins”, this system works in three stages: to see, to think and to do. At the “vision” stage, it is about obtaining data about the situation. This information is of two types: operational data (for example, boiling point) and data from the environment. The next step, which Colin J. Parris conventionally called “think”, is due to the fact that at this stage a “digital double” for different requests can provide options, how best to act in a particular situation or which options are preferable for business purposes. Artificial intelligence uses for analysis, for example, historical information, forecasts on revenues and expenditures, and provides several options that are based on the risks and confidence that these proposals will be able to reduce them. The last step – to “do” – is directly related to the realization of what needs to be done.

With the help of “digital twins”, for example, one can see from within the problems of a physical object.

In production, we do not have to see before ourselves, for example, the whole turbine in order to detect a hole. The technology of “digital twins” will allow to see the problem in real time using computer visualization.

According to Zvi Fejer, executive vice president of software development at Siemens, the digital double is a PLM solution on the road to the transition to Industry 4.0.

What kinds of “digital twins” already exist?

As we said earlier, “digital twins” are actively used in industry: twins-parts (which are built for a specific production unit), twins-products (associated with product release, their main task is to reduce the cost of maintenance), twins-processes their goal may be, for example, increasing the service life), systemic twins (optimization of the entire system as a whole).

According to the research and consulting agency in the field of high technology Gartner, in the near future hundreds of millions of “digital twins” will replace human labor. In some companies this is already being applied. It is not necessary in the state to have an employee who would be involved in the diagnosis of malfunctions in production. In real time with the help of “digital twins” you can get all the necessary data and be ready to repair equipment in advance.

And what about the “digital twin” of the person himself?

For those who want to have a terminator friend who thinks how you helped in everything, was a brother and a friend, we have good news. According to futurist and technologist-theorist, founder of Acceleration Studies Foundation John Smart, this future is already close. He believes that in the near future there will be so-called software agents that will anticipate in advance the wishes, the behavior of their real copy and will perform some actions for their human counterpart.

“Digital twin” will be able to make purchases, make business decisions, engage in social activities – in general, will be able to do everything, for which we sometimes do not have enough time.

Also we can shift all our routine work to our double. In addition, according to John Smart, our digital clones will know our interests, preferences, political views and, if necessary, will be able to defend them, as they will have a more complete historical context and see the modern picture of the world as a whole. And even a feeling of compassion. For example, the “digital twin” will show empathy to us, as it will be able to guess our emotional state.

Are there any downsides to the “digital twins”?

The drawbacks of the “digital twins” are obvious. First of all, the question arises about our security. Digital clones will use all possible resources to replenish information about us. These are algorithms that collect data from social networking accounts, and our personal correspondence, and any documents and files that, in one way or another, concern us. Of course, this can not but alarm: as we have already found out, “digital twins” are able to constantly update and improve. Therefore, one of the first tasks should be the creation of a legal framework for determining the “limits of permissiveness” of artificial intelligence.

However, do not panic about this. Take an example from John Smart – he remains optimistic and believes that “digital twins” can not replace humanity. They will simply become other versions of the person who will be able to coexist peacefully with us: “Our digital systems will become more and more high-quality copies of people. However, this does not mean that “they” will replace “us”, rather they will simply become another version of ourselves. Mankind has been working for many years to put in technology all the best that is in us”.