Smart Factory: More Than the Sum of its Parts

Publish Date

19 SEP 2020

Author

Kai Beckmann

Overview

The use of robots, cloud computing and AI enables more flexible and efficient production in industry. The so-called fourth industrial revolution is the next development stage in industrial production and will represent another significant acceleration of global growth.

What are currently the most important technologies on the path towards a smart factory and where do we stand with respect to this?

Industry 4.0 on the rise

The smart factory is no longer a distant prospect; it has long since moved from theoretical position papers and studies into practical application. And we too have been advancing the modernization of our production for several years. The factory of tomorrow distinguishes itself by combining a whole series of different technologies, which, ideally, mesh together seamlessly. In particular, these technologies include IoT, artificial intelligence, cloud, 5G, and Big Data.

In practice, smart factories can look completely different, depending on the sector and the company. For example, in the Electronics business sector at Merck KGaA, Darmstadt, Germany, a modular system is at the heart of our smart production. As a result of the numerous combination possibilities, the manufacturing processes of thousands of chemicals can be represented with a manageable number of system modules. As a specialty chemical manufacturer with a very broad product portfolio, this gives us an enormous boost in efficiency because we do not need a special production plant for every chemical. And since every module has a digital computing unit, we can control and monitor the entire production at the computer.

Keeping an eye on everything digitally

But what are the next developments? In the factory of tomorrow, all machines, tools and even the products themselves will be equipped with computer chips and sensors so that they can communicate with each other. Thanks to the Internet of Things, every component in the physical factory will have a digital twin. The overall digital image of a factory then enables us to simulate entire production steps virtually. As a result, it would be possible to initially test the settings of a machine for a new product virtually before the machine is actually configured, for example. This saves time and money.

At Merck KGaA, Darmstadt, Germany, among other things, we are currently modernizing our Polyproduction in Darmstadt, where we manufacture liquid crystals, OLEDs and pharmaceutical products, so that it is state of the art. The objective is to network all plant systems digitally and to integrate them in a central information platform, which every employee can access by smartphone. As a result, all colleagues at every site can gain a real-time overview of the entire production process and can respond immediately if there are any malfunctions. Additionally, employees will have the possibility of exchanging relevant information, such as laboratory results or details about raw materials delivery, via smartphone.

Today, robots already carry out production steps in industry fully automatically. Some of these production steps are complex. The interaction between machines and humans will become more and more sophisticated in the factory of tomorrow and the interfaces will continue to be better adapted to our speaking and viewing habits. It is likely that giving robots tasks via voice control, having additional information on individual products in the warehouse shown to us on a tablet via augmented reality or preparing for complex repair work by simulating it with virtual reality glasses will be a matter of course in the not-too-distant future.

The most important raw material of the 21st century

One important prerequisite for managing the enormous amounts of data that arise in a networked factory is a powerful transmission technology. With speeds of up to ten gigabits per second, the mobile communications standard 5G is the most promising prospect here. Data about the smart factory are distributed and stored in the cloud. The software that controls the machines also runs less and less frequently on local systems because computing capabilities can be scaled considerably more flexibly in the cloud.

One particularly smart feature is the ability to use this mountain of data in order to then optimize production processes, for example. Thanks to artificial intelligence, complex data sets can now be analyzed more and more quickly. Big Data analyses can reveal new possibilities for action for companies. These possibilities previously remained hidden among the flood of data. For example, at Merck KGaA, Darmstadt, Germany, we use a software called Comet in chemical production, which allows us to identify deviations in the production process by evaluating huge quantities of data from different sources. The software can thus serve as a basis for the use of AI.

Exploiting the full potential of the digital factory

Preparing the people who will work in the factory of tomorrow for the change through corresponding vocational and advanced training will also be crucial for the success of the smart factory. After all, for them, the digitalization of the factory is a relief on the one hand but also places completely new demands on their own skills and ways of working. Thus, the modernization of our Polyproduction is also accompanied by our own Merck KGaA, Darmstadt, Germany online platform with video tutorials, for example. Our employees can use these video tutorials to find information on how to deal with this new technology.

It’s definitely worth the effort. According to a study by the consultancy Deloitte, companies increased their productivity by 12% on average between 2015 and 2018 through smart factory initiatives. And that is sure to be just the beginning. The full potential of the smart factory will unfold when companies network their entire value chain and all departments, as well as customers and suppliers, grow together in an integrated digital ecosystem. The smart factory will then be more than the sum of its parts.

By {{ authorText(item) }}

{{ item.publicationDate }}

Comments

All fields are required
0 Comments