The Smart Factory of the Future
Editor’s Note: This article was created with expertise from Andreas Dreher, the strategic technology manager at Hirschmann Automation and Control.
When it comes to industrial networking today, many factories and process control facilities around the world are focused on upgrading to managed Ethernet networks. With the long, useful life of industrial devices, there is plenty of old equipment using legacy industrial protocols in active service. Indeed, much of our business involves helping companies upgrade to structured, reliable and easy-to-maintain industrial Ethernet infrastructure.
Even given this reality, it is instructive to step away from current challenges and look ahead to the Factory of the Future. What will industrial production look like 5-20 years from now? What do I need to understand about where factories are going to guide my decisions today? How will my factory compete with brand new factories that use next generation communication systems and concepts? Where does the industrial Internet of Things fit in?
This blog applies not only to discrete manufacturing, but also to automation in the process, energy and transportation industries.
We are fortunate to have insight into this topic from our Hirschmann division, based in Germany where “Industry 4.0” is part of a large publicly funded project, to inform this discussion.
Let’s take a look at what the Smart Factory is and what characterizes its communication systems.
The Smart Factory of the Future will consist of systems that are more intelligent, flexible and dynamic than the ones in use today.
Defining the Smart Factory
The terms “Smart Factory,” “Smart Manufacturing,” “Intelligent Factory” and “Factory of the Future” all describe a vision of what industrial production will look like in the future.
In this vision, the Smart Factory will be much more intelligent, flexible and dynamic.
Manufacturing processes will be organized differently, with entire production chains – from suppliers to logistics to the life cycle management of a product – closely connected across corporate boundaries.
Individual production steps will be seamlessly connected. The processes impacted will include:
- Factory and production planning
- Product development
- Enterprise resource planning (ERP)
- Manufacturing execution systems (MES)
- Control technologies
- Individual sensors and actuators in the field
In a Smart Factory, machinery and equipment will have the ability to improve processes through self-optimization and autonomous decision-making. This is in stark contrast to running fixed program operations, as is the case today.
Figure 1: The Smart Factory of the Future is based on a fourth industrial revolution – Industry 4.0, and is centered on the use of cyber-physical systems.
Key Traits of Future Industrial Networking Solutions
To do this, the future structure of factories will be much different: an inter-connected combination of intelligent production technologies, with the newest high-performance information and communication technologies.
This will provide digitally integrated engineering and horizontal integration across the entire value chain, as well as vertical integration and connectivity across all levels of production.
High-performance, reliable communication technology will exceed what is currently in use. This technology will make it possible to:
- Transfer large amounts of data in real-time and with minimum delay
- Connect a large number of individual devices in a very reliable manner and with the highest standards of data security
- Utilize more and more wireless technologies, both within the plant and for remote connectivity
- Operate in an energy-efficient manner
Sounds wonderful, doesn’t it? It might seem unrealistic to you right now, but I hope that once we break the Smart Factory down to its communication components you will be able to see that it is attainable.
The Structure of Future Industrial Automation Systems
Today’s industrial automation systems consist of several clearly separated levels typically represented as a pyramid with:
- Field level actuators and sensors
- Control level control devices, I/O modules and operator terminals
- A process management level with computers for engineering, supervisory control and data acquisition (SCADA) and MES systems
- An enterprise level with business processes and ERP systems, typically located on servers in the IT data center
Each of these levels is relatively well structured and individual devices can be clearly mapped to one of the levels.
Figure 2: In the Factory of the Future, the field level remains distinct, but other levels migrate to server farms or the cloud.
With Industry 4.0, the system structure changes. The field level remains a separate dedicated level, as it is now, but the devices on it will embed more and more intelligence. As parts of cyber-physical systems, they will autonomously perform many processes. Field level devices will also significantly increase in numbers.
All functions located above the field level will potentially move to high-performance servers located in a server cluster, data center or in a “cloud.” Virtualization, the separation of specific functions and processing hardware, which is already state-of-the-art in the IT world, will become commonplace in the factory.
The advantage of this structure is that it reduces the variety of devices, which results in easier management, better utilization of resources and a clear cost savings.
This approach has not yet been adopted in automation because of issues related to performance, required determinism, reliability, and the lack of fast, low latency communication from the servers to the field level. Nonetheless, these issues will be addressed in new and upcoming systems.
Smart Factory - Examples of Cyber-Physical Systems
Since Industry 4.0 will be built with cyber-physical systems, let’s take a moment to consider what they are. The website Cyber-Physical Systems describes them as:
“….. integrations of computation, networking, and physical processes. Embedded computers and networks monitor and control the physical processes, with feedback loops where physical processes affect computations and vice versa.”
An example of such a system today is the CarTel project at MIT where a fleet of taxis collects real-time traffic information in the Boston area. This information is combined with historical data to calculate the fastest routes for particular times of the day.
Another example that you may be familiar with is the Smart Grid. One definition of it, based on work from the U.S. Department of Energy, is:
“A modernized electrical grid that uses information and communications technology to gather and act on information in an automated fashion … to improve the efficiency, reliability, economics, and sustainability of the production and distribution of electricity.”
Finally, an example for a factory is changing systems so that the energy consumption in a vehicle assembly line is reduced when the line is not operation. Today, many production lines continue running during breaks and weekends. Consider laser welding technology that remains powered up over weekends so it can resume quickly on Monday. This practice consumes up to 12 percent of total energy consumption of the assembly line.
With Industry 4.0 and cyber-physical systems, robots will go into standby mode as a matter of course during short production breaks and power down during longer breaks. Speed-controlled motors that reduce the energy required to run machines will be widespread. Such changes will significantly reduce energy consumption and will be taken into account up front as part of Smart Factory design practices.
Communications in the Factory LAN
Above, I described a vision of manufacturing where systems are more intelligent, flexible and dynamic. In the future machinery and equipment will have the ability to improve processes through self-optimization and autonomous decision-making. This is in contrast to running fixed program operations, as is the case today.
The structure of industrial automation systems will also change. There will still be a separate dedicated field level with actuators and sensors, but in the long term many functions located above that will likely move to high-performance servers located in a server cluster or in a “cloud”.
“Cyber-physical” systems will be important, with feedback loops where physical processes affect operational programs and vice versa. An example of a cyber -physical system is the Smart Grid which aims to improve the reliability and efficiency of the electrical grid system through collecting thousands of data points and acting on them using software management tools.
Let’s now consider the manufacturing LAN and its communication systems. How does it need to change to realize the vision of the Smart Factory?
The Smart Factory of the Future will consist of systems that are more intelligent, flexible and dynamic than the ones in use today. Photo courtesy of BMWblog.com.
Smart Factory - High Numbers of Connected Devices using Industrial Ethernet Protocols
The number of connected devices in a future Smart Factory LAN will clearly be higher than it is today. This is because there’s a need to collect as much real-time data as possible that is relevant to the process. It’s estimated that the quantity of connected devices will double or triple.
The challenge will be connecting this large number of devices at the field level in a simple, cost-efficient manner. Of course demanding requirements for performance and reliability will still need to be met.
The use of field busses will decrease significantly to make way for consistent and unified communication via an Ethernet network. All communication will be based on IP protocol families and Ethernet will be the underlying communication protocol, regardless of whether the connection is wired or wireless.
Smart Factory - Increase in Use of Star Network Topology
As is the best practice today, future networks with high numbers of devices should be hierarchical to simplify network management and operation. The field level should be segmented into manageable communication cells, such as by production units, or any other logical or physical units. The difference will be that the amount of data generated in the cells will be significantly higher than it is today.
The network will still use star, line or ring topologies, or a mix. The use of star topologies will increase, however, because they have some advantages – such as lower latency and higher reliability – compared to other topologies.
The disadvantage of a star topology is that the failure of a switch will disconnect all attached devices. Nonetheless, simulations clearly show that one larger switch has a higher total reliability – more precisely, a higher Mean-Time-Between Failure (MTBF) – compared to a system consisting of many cascaded, small switches. This is the reason why star topologies are used in data centers today.
Lines or rings will be used too, because certain topologies might have advantages in cabling. Additionally, the use of more complex structures, such as extensively meshed network topologies, will increase. With the adoption of new protocols, these networks will need less management efforts.
The Smart Factory of the future will use network topologies familiar to us today, but with higher numbers of connected devices delivering higher amounts of data.
Will the Smart Factory be Wired or Wireless?
In the future, will all devices be connected by cables and wires or will everything be wireless? In the industrial applications of the past, communications were almost exclusively based on wired networks.
In recent years, however, wireless systems have found increasing use. They have been adopted most often for non-critical industrial applications, such as configuration and monitoring, transfer of peripheral data and for mobile worker applications.
The challenge with radio is that it is a “shared media,” i.e., all devices share a certain frequency range. If a device is transmitting, the channel is busy. Radio communication can also be error prone. Other radio systems, other electromagnetic influences or objects can affect transmission and significantly deteriorate quality, bandwidth and latency.
The sporadic loss of data packets is the norm in some radio systems and has to be handled by the applications. This is done at the expense of throughput and latency. While this may be acceptable in enterprise wireless deployment environments (like in offices and businesses), industrial wireless products need to be designed from the ground up for reliable performance.
Well-designed industrial wireless products are now employing techniques like:
- Enhanced electrostatic discharge (ESD) protection for hazardous environments
- Wireless mesh technology for quick network reconfiguration and service assurance,
- Redundancy protocols like Parallel Redundancy Protocol (PRP) for wireless communications.
These intelligent technologies help industrial wireless applications adapt to radio channel performance issues and deliver much more dependable systems.
Nonetheless, reliability requirements will drive the choice of communication technology, wired or wireless, in the Smart Factory. Significant use of wired communications can be expected, but the flexibility of deployment of wireless connectivity will drive increasing usage of suitable industrial wireless products.
Communications in the Smart Factory LAN
What will industrial production and automation systems look like 10, 20 and 50 years from now? What do I need to understand about the communications requirements of the factory of the future to guide my planning today?
Part of the vision of the Smart Factory of the Future / Industrial Internet of Things is that all relevant data will be available in real-time, leading to faster and smarter decisions. This in turn will lead to the design of more flexible and efficient processes.
The full vision is described at the beginning on this topic. I then covered how network topology and the balance between wired / wireless applications will change.
Now, the final installment looks at the data rates, cyber security and reliability characteristics of the industrial facilities of tomorrow.
The sought-after business benefits of the Industrial Internet of Things (IIoT) are driving technology innovation that will lead to the Smart Factory of the Future becoming a reality.
Smart Factory Data Rates: Gigabit Ethernet is Standard
Needless to say, the future Smart Factory will communicate using faster rates of signal transmission. Today, Fast Ethernet with 100 megabits per second (mbps) is the standard for industrial applications.
In the IT world, Gigabit Ethernet (1000 mbps) has been state-of-the-art for quite some time. Most PCs today support this high data speed. Even if Fast Ethernet is good enough for the amount of data produced by an automation device, the trend will be to use Gigabit in the near future.
Fortunately new chips are already integrating Gigabit Ethernet interfaces, thus decreasing the cost for a faster connection. Plus, advances in semiconductor processes also will lead to lower power consumption, so that today’s price and power consumption arguments will soon be irrelevant.
Along with wired network speeds, wireless network speeds are also increasing. New WLAN technologies, like IEEE 802.11ac and .11ad, are enabling wireless to quickly close the performance gap with wired communications. Such technologies are being perfected now in enterprise deployments. Their adoption in the Smart Factory is expected over time.
Smart Factory Security: Secure-by-Design Devices
The downside of increasing connectivity and use of open standards in industrial networks is a significantly higher risk of cyber security incidents. These include deliberate attacks (estimated to account for about 20% of incidents) as well as unintentional human errors and device conflicts.
The Smart Factory network will need to support security functions, including:
- Encryption to ensure the confidentiality of the data and prevent any unauthorized interception of data, particularly important for data traffic running over public networks.
- Access control to ensure that only devices allowed to communicate with each other can do so, to prevent unauthorized access during operation.
- Authentication as another element of access control to block unauthorized devices and users.
They will also include a security chain that can be built by devices from the hardware and firmware up to the applications. This will help ensure that each component in the system – software, connection and transaction – is trustworthy, safe and secure. While this sounds like wishful thinking compared to the ICS cyber security practices in place today, vendor-neutral organizations like the Trusted Computing Group are working on standards to achieve it.
Other security measures include the detailed logging of all events and changes via log files to precisely track network activity. Network management and security tools can be used to monitor the network behavior and traffic. They can also detect potential threats, like abnormal traffic patterns or unauthorized access attempts, and take appropriate countermeasures.
Reliability Requires Multiple Types of Redundancy
One aspect of reliability in a Smart Factory system is network redundancy, or the behavior of the network in the event of a failure. Disturbances and interruptions in the communications network can never be completely avoided. Failure of a cable or connector due to mechanical overload, the failure of a power supply unit, or even short- term shutdowns for maintenance reasons can affect network traffic.
To ensure that only the smallest possible part of the system is affected, redundancy capabilities that redirect traffic to an alternative path are needed. Redundancy protocols ensure that there is only one logical path between any two devices, even if there are multiple physical paths. Only one of the paths must be active and transfer data, while the other paths are in stand-by mode.
There are a number of protocols on the market that differ both in the switchover time and the supported topology. These include:
Rapid Spanning Tree Protocol (RSTP)
Media Redundancy Protocol (MRP)
Parallel Redundancy Protocol (PRP) and High Availability Seamless Ring (HSR)
PRP: Two independent networks
HSR: One ring network
Multiple redundancy protocols will contribute to high reliability in the Smart Factory of the Future.
There are also other approaches that are currently available or in the works, such as:
- A distributed link aggregation protocol (Distributed Resilient Network Interconnect)
- The Shortest Path Bridging (SPB) protocol (IEEE 802.1aq)
For Smart Factory applications, the required network redundancies must be analyzed carefully before a protocol is chosen. Often, there will be a mix of network segments that use full redundancy based on PRP. For other segments, RSTP, MRP or a distributed link aggregation protocol will be the best choice to achieve network reliability.
IoT Technology Innovation Will Make the Smart Factory a Reality
The success of the Smart Factory / IIoT vision largely depends on the underlying communication technologies achieving high performance levels. If the communication infrastructure cannot meet the demanding requirements, many applications will not work as intended.
Currently, there are many ongoing efforts to close the remaining gaps with new, innovative enhancements in data communication technologies. There are several challenges to overcome, but from today’s perspective, it will one day be possible to provide all the necessary elements to make the Smart Factory vision a reality.
Editor’s Note: This article was created with expertise from Dr. Tobias Heer and Dr. Oliver Kleineberg from our Hirschmann industrial networking group.
*The restrictions are not related to the total number of switches but are related to the diameter of the network.