Main image of CPS: A Key Technology for Realizing Autonomous Smart Factories

Smart Factories: Giving Form to Digital Transformation in the Manufacturing Industry

CPS: A Key Technology for Realizing Autonomous Smart Factories

Many companies in the manufacturing industry around the world have begun work on building smart factories. A smart factory is a future factory aiming for solutions to issues faced on the frontline, improvements in productivity and quality, and other features by proactively utilizing artificial intelligence (AI), IoT, and other advanced digital technologies. We give an explanation here about the cyber-physical system (CPS) – the future form of the production system occupying the core of smart factories (Fig. 1).

Image 1 of CPS: A Key Technology for Realizing Autonomous Smart Factories
Fig. 1: CPS image and function configuration

Differences between conventional FA and smart factories

Productivity and quality have been improved by utilizing factory automation (FA) in many factories up to now. Various manufacturing systems and industrial robots that automate processes in production sites have been lined up to mass produce high-quality products with high productivity. 

What are the differences in the purposes and techniques between factories that have implemented this conventional FA and smart factories being tackled by many companies now (Fig. 2)?

Image 2 of CPS: A Key Technology for Realizing Autonomous Smart Factories
Fig. 2: Differences between factories that have implemented FA and smart factories

Factories that have implemented FA are improving productivity and quality by comprehensively automating a series of processes including processing, transportation, and inspections on production lines. However, the lines that are built are not flexible with respect to fluctuations in production items or production volume. That is because the line structure, the systems and equipment specifications, and their operating conditions have been optimized based on a predetermined workflow and predetermined production conditions.

In contrast to this, smart factories are built on the assumption that they will respond promptly and appropriately to changes in production items, increases and decreases in the production volume, deviations in the production conditions that occur during operation, malfunctions and failures of systems and equipment, and other elements. These changes and unexpected situations can occur on a daily basis in factories in any industry. The ability to respond to these is directly linked to the value of the factory and, by extension, the competitiveness of the company in the manufacturing industry.

CPS: Looking into the future a little ahead of time on the production line

The CPS is an information system that plays the role of the brain and the chain of command in smart factories. Put simply, the CPS is a system that predicts what will happen on the production line in the future a little ahead of time. Its mechanism and functions are as follows.

First, a computer model that has the same functions and performance as the production line itself or the systems and equipment placed in it is prepared in a virtual space. The movements and conditions are then reproduced by collecting operating data of the systems on the line in real time with sensors and incorporating that into the model. This model that reproduces the same behavior as the actual systems is called a "digital twin."

It is possible to attempt various operating conditions in the digital twin to reduce risk and find the optimal solution in advance even when changing production items or the production volume with a CPS. Furthermore, it makes it possible to sense when on-site systems will fail in advance and to then take measures against that ahead of time.

CPS produces further diverse value with an increase in sophistication of IoT

Highly experienced engineers familiar with production sites played the role to optimally operate factories until now. In other words, it can be said that the operation of factories relied on individual experience and intuition. 

Utilizing the CPS makes it possible to provide a clear basis for decision-making based on data to engineers who lack experience. It may be possible to operate factories autonomously with a CPS alone if the accuracy of looking into the future improves. In addition, if an IoT system is built that can grasp phenomena that occur in factories without overlooking them by placing sensors in locations previously out of reach of the human eye, it will be possible to expect higher productivity and quality than ever before.

Moreover, it may be possible to expand high-added-value business called "mass customization" to make custom products with specifications that meet the requirements of each user. It takes a surprising amount of time and effort to make products with different specifications. However, it may be possible to automate this with a CPS to be able to start custom production without constraint.

Evolution of sensors, actuators, batteries, and other parts is essential for the evolution of CPS

The digital twin is placed in a data center with powerful computing power in most CPS. However, it is necessary to enhance the performance of three technological elements in factories to realize highly accurate digital twins (Fig. 3).

Image 3 of CPS: A Key Technology for Realizing Autonomous Smart Factories
Fig. 3: Devices that influence the CPS system configuration and the performance of each element

The first element is edge devices that collect data from the production site and then feed back the analysis results to that site. It is necessary here to have sensors that detect data that reflect system movements and conditions such as vibrations, temperature, sound and video, actuators that control the movement of systems, and other elements. It is essential that these electronic parts have high reliability to move accurately even under harsh environments in factories. Moreover, there is also a need to reduce the power consumption of devices and to increase the capacity of devices to lengthen their operating time.

The second element is edge computers that process information in locations close to the production line. A microcomputer is used here to screen and simply process the collected data. The needs to seek advanced judgments by embedded AI are also growing. The performance of microcomputers will advance and the capacity and speed of storage will no doubt increase in the future.

The third element is close-distance networks that connect edge devices and edge computers. EtherCAT*1, CC-Link*2 and other highly reliable industrial networks are used as wired networks. High-quality noise suppression products are important here. Bluetooth and Wi-Fi have higher reliability as wireless networks. If long-distance wireless communications are required, the use of Low Power Wide Area (LPWA) and 5G is also considered. It is essential to have parts that realize the wireless element of these in a small size with low power consumption and high reliability.

*1: EtherCAT is an industrial network standard developed by Beckhoff Automation in Germany. It is compatible with the Ethernet that is used in the construction of FA systems. The functional requirements, authentication procedures, and other elements are defined and managed by the EtherCAT Technology Group (ETG).

*2: CC-Link is an industrial network standard developed by Mitsubishi Electric. The initial standard was just technology with unique specifications optimized for use in factories and other production sites. The Ethernet-based standard called "CC-Link IE" has also appeared and is being used with the aim of linking up to the core information systems and production systems of companies in recent years. The CC-Link Partner Association is conducting activities to spread and promote this standard.

Other Links

Related articles