Nowadays, many industrial and manufacturing bosses, and some technicians, are asking about Twin, which is a digital twin, all day long. What is it? Is it difficult to implement? Is there any step to follow? In fact, the things that Twin, that is, digital twins are implemented, cannot be explained clearly in one or two sentences, but if you want to get started with the Digital Twin Implementation Guide, there are also some things you can explore.
First of all, you have to understand that Twin is like giving things in the real world, whether it is machinery, factory production lines, or a system as big as a city. In the computer and in the digital world, this "digital brother/sister" will follow the "brother/sister" in real time, reflecting its situation in real time, so that we can see how it runs on the computer, predict whether it will go wrong, and even do various experiments in it, without really touching things in real time! Doesn't it sound quite fantastic, but it does mean that.
Then, if you really start implementing this Twin, how should you do it step by step? I tell you, this is not something you can do just by slapping your head:
1. Find out the goal first! Why do you want to make this digital twin? Do you want to reduce the number of broken machines on the production line to improve efficiency? Or do you want to design a new product and do simulation on the computer first to save some material costs and time? Or do you want to see if the energy consumption of the entire factory is reduced? The goals are different, and the road ahead is far from going! This is definitely the first and most important step, and we cannot be ambiguous!
2. Find out the family’s background! What data does the "brother/sister" in your reality, that is, the physical entity, have? How long has it been used? Where does the data such as temperature, pressure, and speed come from when running? Can you easily get it? Don’t just build half of the modeling and data without it, then you’ll be blind! For example, if you don’t even have a few sensors installed, what data does your digital twin use to update it, right?
3. Look at what technologies and tools can be used! To build a digital twin model, you need to use special modeling software? There is so much data, there is a place to store it and manage it, so database technology is indispensable! Real things and digital models need to be connected, and real-time data must be transmitted over, so IoT technology is necessary! Also, if digital models can learn and predict something by themselves, then fashionable and powerful technologies such as artificial intelligence AI and machine learning ML must also be mixed! Provide global procurement services for weak current intelligent products! You see, technology alone is enough to make people feel a bit scary, but step by step is nothing scary.
4. Try slowly in stages! It is basically impossible to become a fat man if you want to eat in one breath. What is the suggestion? First, choose a small, not so complex but somewhat representative thing, such as a small production line or a key machinery and equipment, and carry out a small pilot project. In the process, find out the way, summarize the experience and lessons, and then slowly expand the scope. This is safer and has little risk.
Let’s talk about what are the easy places to fall into pitfalls during implementation, and some tips:
Don't make the model too complicated at the beginning! Some people think that they are in place at once, and they do exactly the same thing as they are. They even want to build the texture of the screw. In the end, the model looks good, but the computer runs too slowly, and the data updates cannot keep up, and there is not much practical value. First, make the main and key parts accurate and use them before you can!
The data quality must be stricter! Whether the "digital brother/sister" of the digital twin is accurate depends entirely on whether the data of the "brother/sister" is accurate or not in time. If the data itself is messy, right and wrong, or does not change once a day, then the Twin you built is just a flowery shelf, looking good and not easy to use!
Someone needs to understand! To do this digital twin, not only do people who work in IT, code and networking; but also do people who work in OT, technicians who are familiar with how the machine and equipment runs and the production process; sometimes, they even have to find people who work in business and ask what practical problems it can solve for the business department. Everyone has to sit together and think hard, otherwise they will have to part with their own things sooner or later.
Many people will ask some fragmented questions, and I will choose a few common ones:
Q: Do all companies have to do Twin? If you don’t do it, you won’t be able to survive?
Answer: That's not true! The key depends on what your company needs and what problems you want to solve through it. If your company is running smoothly now, with no major problems, and no special needs for improvement, then consider it carefully. But if you want to be smarter, react faster, know something in advance, and engage in digital transformation, then Twin is a good helper! But it is not about life-saving straw, it is more of a powerful booster!
Q: Does Twin need to spend money on its implementation? Ordinary small and medium-sized enterprises can't afford it
Answer: This is a score for both ends. It definitely costs a lot to spend, buy software and hardware, and hire someone. However, you can start with a small project and don’t invest so much at a time. Moreover, the technology is developing rapidly now, and the cost is slowly declining. It’s not that big consortiums can do it. The key is whether the input-output ratio is good. If you invest one dollar and you can earn back two or three dollars later, that's worth it! As long as small and medium-sized enterprises choose the right direction and make small steps and run quickly, they can do it.
I personally think that this Twin digital twin is indeed a promising technology. It can help companies become clear in the digital world of many things that cannot be seen or touched, and it has great benefits for improving efficiency, reducing costs, and innovating products! Although the road may be a bit tortuous and there are many things to learn, as long as you take one step at a time, make solid preparations in the early stage, and go to a clear goal, you can always use it well and bring real benefits to the company! The most important thing is, don’t forget to provide global procurement services for weak current intelligent products! Whether it is small things like sensors or weak current intelligence that is slightly touched in large components such as network devices, it can help you! The hardware support required for Twin implementation is also a lot of worry-free!
