The term Industry 4.0 implies more than the next step in industrial automation—it is shorthand for the fourth stage of the industrial revolution and is a prediction, not a description of an actual state of affairs.
An article in the April 2011 issue of the newsletter of the Association of German Engineers (VDI), suggests that the first industrial revolution began in the 18th century with the introduction of machinery to production. The second revolution entailed “the division of labor and the mass production of goods with the help of electrical energy,” which extended from the start of the 20th century until the mid 1970s. The third revolution was the introduction of electronics, computers and IT to automate industrial processes.
We now may be entering the fourth revolution, in line with a concept that originated in Germany, where the government is supporting “The Industry 4.0 Project” with a budget of up to €200 million (nearly $250 million).
Mark Watson, associate director for Discrete and Process Automation, IHS Technology, writing in the IHS Quarterly/Technology Q4-2014 article, “Smart Manufacturing Gains Momentum,” uses the term “smart manufacturing.” He describes it as “a progressive, forward-thinking concept perfect for the unfolding millennium—of multiple manufacturing companies and their suppliers, customers and partners, all linked together via leading-edge industrial protocols and networking, able to make astute decisions and adapt to rapidly changing conditions in factories as diverse events occur.”
(Watch Tim Dawson, senior director, Industrial Automation for IHS Technology discuss "Industry 4.0: Opportunities and challenges for smart manufacturing.")
Professor Dieter Wegener, head of the Industry 4.0 Office at Siemens, cautions, however, that “it could take 15 years for all of this to happen.”
As explained in the VDI article, the concept was initially developed to define a path for German firms to compete in global manufacturing even though they are a “high-wage region.” Another motivation, according to Watson, is that the current expert workforce, predominately composed of baby boomers, is leaving the workforce and not being replaced in adequate numbers.
The economic scale of the markets involved with industrial automation was $169 billion in 2013 and was forecast to reach $179 billion in 2014 and $210 billion in 2018, according to IHS. The exact nature of these markets is, therefore, of major consequence for businesses, both to market goods such as motors, generators and controllers, and also to automate operations.
“The end result is a complex but vibrant ecosystem of self-regulating machines and sites, able to customize output, allocate resources optimally and offer a seamless interface between the physical and virtual worlds of construction, assembly and production,” says Watson. The task over time is to develop specific tasks and methodologies that will move toward these goals.
A smart factory is one in which the processes of planning, design, manufacturing, procurement and analysis are linked in a common network of communication and control. The Industrial Internet of Things (IIoT) is necessary, but not sufficient, for a smart factory. IIoT will provide the plant floor infrastructure for Industry 4.0—an industrial-hardened array of smart sensors, controllers and HMIs—and the networking technology to integrate them.
“Industry 4.0 requires distributed intelligent systems, physical systems and virtual digital data to merge into cyber physical systems,” says Tom Burke, OPC Foundation president and executive director. OPC is the interoperability standard for the secure and reliable exchange of data in the industrial automation space and in other industries.
Industrial Ethernet offers what some consider to be the ideal networking infrastructure for Industry 4.0. It enables seamless integration between the manufacturing and the enterprise levels, since business applications have traditionally been connected by Ethernet. The key is that each device connected to the Ethernet network has a unique IP address. This means that each of them can be programmed to perform and communicate as needed by the overall system.
There also has to be a supervisory system to coordinate and control the cyber physical systems and make sense of all of the available information they provide. Enterprise resource planning (ERP) software “supports an organization’s entire business process,” according to Enterprise Resource Planning Software (ERP) Information.
It integrates all functional departments, for example supply chain management, accounting, inventory control and manufacturing. Manufacturing execution system software (MES) is used by plant managers and production personnel to manage and monitor work-in progress on the factory floor—to “support collaborative manufacturing strategies that are designed to integrate disparate data streams from a company’s supply chain, factory floor and ERP system.”
The ERP system also can act as a gatekeeper to supervise communications with the world outside of the enterprise. Systems can become virtual segments of the enterprise network by communicating via Internet, wide-area networks (WANs) and cellular networks. This enables remote data centers (the “cloud”) to be used for dealing with with the huge amounts of data accumulated by the integrated networks. Enterprise logistics operations can also be coordinated over Internet or cellular networks.
“Smart sensors can detect and report low inventory to the company’s enterprise resource planning (ERP) or to a human decision-maker, resulting in a refill order," says Watson. For example, a sensor might report a delivery van’s breakdown and order a second van to transfer supplies as well as a tow truck to rescue the disabled vehicle.
Data vs Information
Distinguishing between data and information becomes important in discussing Industry 4.0. Information is composed of individual data points, for example, the temperature of a liquid. Information based on this data could be combined with other data such as a time base to indicate a trend: How has the temperature changed over time, at what rate and for how long?
One question is where and how should this conversion of data into information take place. There are a variety of answers. For example, a temperature sensor with an embedded microprocessor can calculate trends and alarm conditions on its own and send this information to the PLC that’s controlling the process. That PLC can, in turn, transmit the alarm over a wired network to an HMI in a control room. At the same time, it also can broadcast the alwarm wirelessly to key personnel who are responsible for that process and who carry devices with HMI screens. Transmitting specific information rather than masses of data points reduces the network's required bandwidth as well as the amount of data that must be stored.
On the other hand, many advantages exist to transmitting raw data. The virtue of analytics is that the same data can be used for different kinds of information. The amounts of stored data can be converted into useful information by analytics software.
Data points can be thought of as pixels in a digital image: the more pixels, the clearer the image. In the same way, the more data points, the more accurate the information that can be built from them. It would, therefore, seem to be advantageous to use as much of the data as possible. Transforming data into useful information can be controlled by software that sorts the data points to answer a specific kind of question. Different algorithms can be applied to the same data set to answer completely different questions.
Information for Analysis
For example, every time a widget comes to the end of a manufacturing assembly line one data point might be sent out on the Ethernet network. Each discrete point will be time stamped, stored and totaled on a server. These elementary pieces of information could be used throughout the factory for different purposes.
If the production rate slowed, for example, the controller might automatically call up data from object sensors earlier in the process to locate precisely where the slowdown occurred. It could also call up speed information from motor controllers. These two pieces of information could be used to decide whether the cause was a defective motor controller or a mechanical problem with the conveyor.
In addition, the controller also could call up data from temperature sensors embedded in the motor driving the conveyor to decide whether the motor is in the early stages of failure. Depending upon the readings, an alarm could be sent to a floor supervisor and/or a maintenance technician to check into the problem and take appropriate action.
The system also could attempt to self-correct by slowing the motor controllers before sending an alarm. The data could be fed to a different algorithm to determine the production rate and decide when it’s time to reorder parts. That information could be fed to a purchasing agent or it could automatically generate a purchase order, which would be sent electronically to the supplier. In this way, the rate of the motor winding's temperature rise could generate a warning that the motor needs maintenance so that replacement parts can be available.
New developments are enabling huge amounts of data to be stored and manipulated with a relatively small investment in technical infrastructure compared to older enterprise-level database management systems and hardware. This can be accomplished by an array of low-priced computers and hard drives, coordinated by open-source software such as Hadoop or Spark. This architecture has the flexibility to expand or contract as system requirements change.
Will Industry 4.0 Happen?
The difficulty in predicting the future of Industry 4.0 is that the phrase itself is only a few years old, yet it has grown to encompass a wide array of concepts. Industry 4.0 is a moving target that likely will evolve as more pieces of it are implemented and evaluated.
“Even if facilities and machines will one day be able to organize themselves much more independently than they do today," says Siemens' Dieter Wegener. "Production processes will still have to be managed at a higher level in order to, for instance, set cost targets, deadlines, environmental goals and other objectives.”
Questions or comments on this story? Contact an editor: firstname.lastname@example.org