The Internet of Things (IoT) promises more revenue, higher efficiency and lower costs, and enables new business models. Many companies are reluctant to take the step towards digitization, however. For every second business a lack of IoT standards is a hurdle, as shown by the VDE (association of electrical engineering, electronics and information technology) trend report 2016 "Internet of Things / Industry 4.0".
"Like a telephone call in China"
One would think that with the Internet Protocol, it is also possible to have a global end-to-end communication in the Internet of Things. This is largely true on the network level of the so-called OSI layer model
. "But this is only as helpful as a telephone call in China when you don't speak Chinese," says Professor Axel Sikora, Scientific Director at the Institute of reliable embedded systems and communication electronics (ivESK) at Offenburg University of Applied Sciences. "As long as the machines speak different languages on the application level, they won't understand each other despite the existing connection."
In the meantime, some frequently used protocols are being used on the application layer at least for connecting machines to cloud services. These include in particular MQTT
(Message Queue Telemetry Transport) for exchanging messages between devices, LWM2M
(Lightweight Machine-to-Machine) for IoT device management, and OPC UA
(Open Platform Communication Unified Architecture) for machine-to-machine communication.
The OPC standard already currently permits various machines in industrial plants to talk to one another. The OPC HDA specification, for example, is used to transmit history data, OPC DA serves to send real-time data, and OPC AE can be used to transmit alarms and events. However, three different applications are also required for these requirements.
OPC UA, as a unifying protocol stack between the different layers, uses open, platform-independent protocols, which support – besides PCs – also cloud servers, as well as Internet standards such as TCP/IP and, in addition to Microsoft Windows, also other operating systems such as Google's Android, Apple's iOS or Linux and Unix. "Over the last few years, OPC UA has developed an outstanding position, especially in the area of industrial automation, and will surely be one of the players who will influence events over the years or even decades to come," Sikora predicts.
Real-time control and synchronization of machines
What OPC UA still cannot offer at the moment is real-time communication, which requires the lowest possible latencies. To achieve this, several industrial companies, including ABB
, Bosch, Cisco, General Electric and Kuka, are working on the OPC UA extension TSN (Time-Sensitive Networking
) as a new standard that is also real-time capable. The manufacturers want to support this uniform communication protocol in their future products to enable real-time control and synchronization of machines over a standard Ethernet network. According to Sikora, although Ethernet standardizes the physical layer in industrial automation, there are still numerous industrial Ethernet variants on the data link layer. "Perhaps TSN will succeed in removing this – sometimes unnecessary – variety."
Don't back the wrong horse when it comes to IoT standards
ho actually sets the standards? The specifications for OPC UA TSN, for example, were developed by the OPC Foundation
in collaboration with the Industrial Internet Consortium (IIC). "If you consider the processes of standardization over the past few years, however, it is clear that the official standardization bodies are increasingly losing their role as trendsetters, and are more and more assuming only what has been envisioned in industrial consortia, alliances, or more loosely organized ecosystems," says Sikora. Another trend has only been apparent for a few years now: namely that very large companies such as Amazon, Google or Samsung "do not even form consortia anymore, but push their solutions directly into the market."
This could be a potential impasse for companies that are planning to step into the Internet of Things. When it comes to implementing digitization, it is important to use and promote developments that are as open as possible," says Sikora. "Unfortunately, some players, especially large ones, in this market continue to try to bind customers to their closed products. In most cases, the barriers to entry are particularly low. But this inexpensive entry can be very expensive in hindsight if you have backed the wrong horse."
On-site or in the cloud – an important decision
If you make the right considerations before connecting your factory
to the Internet of Things, for example, the business potential is considerable. Companies should define beforehand what should be in the cloud and what should not. "Everything that can be exchanged locally or regionally should also be retained there," Sikora recommends. Companies could certainly provide many simpler services locally, such as data storages, for example. This also has something to do with the security-related principle of data minimization, which applies equally to private individuals and corporate information.
However, if devices and machines from different manufacturers are to understand each other across an entire supply chain, the cloud can play to its strengths. Public infrastructures always bring benefits when data is to be exchanged across locations or manufacturers and extensive services are to be used, or when redundancy is an important issue. IoT platforms from the cloud have appropriate interfaces to connect the machines and provide appropriate bridges or middleware. "A key element is also that the cloud forces the machines to use a popular language," says Sikora. "However, one cannot overlook the fact that the translation between different cloud providers is still difficult."
The principle of learning by doing: keep up with the times and time won't pass you by
An overriding, unifying IoT standard across all layers is yet to be found – but this should not deter companies from developing an IoT strategy right now. "It's important to secure an appropriate place in the value-added chain as early as possible before others do so," says Sikora. Learning by doing and learning from mistakes are important and often productive in this innovative field.