The statement by Michael Volz, HMS

We live in a world in which rapid technological change is part of everyday life. It is therefore reasonable to assume that Industrial Ethernet will soon retire the fieldbus. However, the fieldbus is still with us several years after the introduction of Industrial Ethernet and is even gaining in popularity, albeit admittedly at a slower rate than Ethernet.

Michael Volz, Managing Director of HMS Industrial Networks Germany, relies on 'peaceful coexistence': "Gateways can be used to integrate components such as robots, sensors and controllers into various industrial networks.

© HMS Industrial Networks

Our company's analyses and a comparison with several renowned market reports show a growth rate of 17% per year for Industrial Ethernet installations, which now account for a third of all industrial networks. Fieldbus networks, on the other hand, still make up the majority of installations and are currently growing by 7% per year. Both rates are well above the growth rate of automation components in general.

Today, components such as robots, sensors and controllers can be integrated into various industrial networks via gateways, both on a fieldbus and Ethernet basis. We are also seeing little sign of network consolidation: CC-Link, Profibus and Modbus are all still popular, although they vary from region to region. The decision for the fieldbus used in each case often seems to be predetermined by the previous selection of a PLC, so that the developer is not free in his choice.

Fieldbuses are excellent for transferring I/O data and in some cases they seem to be preferred over Ethernet solutions because their installation, configuration and maintenance are already known. Thanks to their low level of complexity, they are easy to install and very reliable. Developers are therefore not lazy or conservative, but choose the best solution for the specific task.

System developers and users naturally always want to stick with existing systems and adapt them to new requirements. They cannot be expected to write off their investment in networks at field level without good reason. This applies equally to system operators and manufacturers/machine builders in the field of series production.

However, they are happy to make the switch to Ethernet if it offers clear advantages, such as better processing of process and IT data, greater functionality, flexibility and speed, as well as easier integration into higher level networks such as the control level and planning.

Our experience shows that fieldbuses are still attractive, but also that the networking of components in production environments is increasing. Until a few years ago, many factories refrained from fully integrating all processes. Instead, some machines or processes were automated, whereby these 'automation islands' could barely communicate with each other, if at all. Nowadays, the trend is towards full integration of all components and, in addition, connection to enterprise management systems. Terms such as 'Machine to Machine Communications', 'Industrial Internet of Things' and 'Industry 4.0' have become established.

We can now see that the transition to Industrial Ethernet will be a long-term process. It is practically impossible to predict when Ethernet will dominate. At the very least, it is likely that fieldbuses will still have their place in five years' time.

Organizations that support technologies for industrial networks will continue to support fieldbuses for the foreseeable future, despite the advancement of Ethernet. For example, CLPA has ensured that its Ethernet technology (CC-Link IE) offers significant technical advantages over its fieldbus options, most notably industry-leading gigabit capacity, which enables high and extremely fast data throughput. As in the future not only individual production machines and processes will be connected to each other, but also to the internal business systems and via these to the systems of the distribution network and suppliers, gigabit capacity and Ethernet capability will not be a luxury in the future, but an absolute necessity.

The bottom line is this: Even though the market for industrial networks is clearly moving in the direction of Industrial Ethernet, it will be many years before we say goodbye to fieldbuses for good. For the foreseeable future, fieldbus and Industrial Ethernet will coexist peacefully.

The statement by Heinrich Munz, Kuka

In the networking discussion, a distinction must first be made between the standard - i.e. IT Ethernet - and the fieldbus-based Ethernet variants generally grouped together under 'Industrial Ethernet' - Profinet, Ethernet/IP, Ethercat, Powerlink, CC-Link IE, Sercos 3. It is indeed true that standard IT Ethernet does not yet meet the real-time capability of fieldbuses and therefore also the real-time capability of Industrial Ethernet and therefore cannot yet replace them.

However, all old and new fieldbus technologies - whether based on Ethernet or not - cannot fulfill the following exemplary requirements for modern automation in the sense of Industrial IoT or Industry 4.0:

• Zero Configuration, also known as Zero Touch: For flexible automation or the use of wireless devices on a large scale, it is not acceptable to have to configure their communication bases before commissioning the devices. However, we have this situation in all fieldbus landscapes in the automation industry today. Before commissioning, major configuration orgies must first be overcome by qualified fieldbus experts. Dynamic registration/deregistration of ad-hoc communication relationships is only possible via crutch solutions called HotPlug or Hot-Connect, whereby the communication relationship that may be entered into later must already be known and planned in advance.
• Cross-manufacturer interoperability: It must be possible for devices from all manufacturers to communicate directly with each other interoperably and without gateways or similar money- and performance-consuming makeshift devices, following the example of the cell phone industry. Today, this is only possible within the respective fieldbus technologies of the driving companies, which want to create the impression of openness through house and yard associations and thus create the necessary community and keep users in the technology trap. Communication from fieldbus group A to fieldbus group B is only possible via gateways or similar.

For these reasons, there will initially have to be another solution in addition to the conventional fieldbuses for an automation landscape characterized by Industry 4.0. This solution is emerging with OPC UA and the expansion of the IEEE 802.1 Ethernet standard to include real-time capability, which is currently being standardized under the keyword TSN (Time Sensitive Networks). In addition to deterministic data transmission in the sub-millisecond range, the main objectives of the new TSN standard are the two aforementioned requirements of zero config and cross-vendor interoperability.

The growth rates for Industrial Ethernet (17%) and fieldbuses (7%) that have been mentioned so far do not, of course, take into account the upcoming introduction of Time Sensitive Networks (TSN), as TSN does not yet exist. So from the user's point of view, we are still talking about a comparison between plague and cholera for the fieldbus and Ethernet variants.

What's more, from the user's point of view, the necessary consolidation of networks is not taking place. This is due to the fact that technologies that are introduced and reasonably successful in our industry remain in place for several decades. For this reason alone, we will have to deal with the legacy of conventional fieldbuses for many years to come, even after the widespread introduction of OPC UA and TSN. That is why the aim of OPC UA, TSN or Industry 4.0 is not to replace conventional fieldbuses from one day to the next, but initially to create a way of operating manufacturer-independent M2M communication - including real-time capable - in addition to the existing proprietary solutions.

The RAMI 4.0 model of the Industry 4.0 platform takes this into account through the so-called integration layer, on which the conventional fieldbuses are located, and the communication layer, which exclusively provides OPC UA for M2M communication. Gradually, OPC UA with TSN can replace one or two applications of conventional fieldbuses. The system integrator or plant operator will decide when this happens.

The ETG image shows the path that lies ahead for all fieldbuses in an Industry 4.0 scenario: Northbound into the production network, manufacturer-independent OPC UA. Southbound into the machine or local field, conventional fieldbuses.

© Ethercat Technology Group

At SPS IPC Drives 2015, four fieldbus associations announced that they would be creating an OPC UA companion specification for their proprietary fieldbuses. This makes it possible for the data models and profiles of the fieldbuses - in which a great deal of money and effort has also already been invested - to live on at OPC UA level. The Ethercat Technology Group (ETG) has published an apt illustration of this, according to which Ethercat EDP or EAP is used 'downwards' (southbound) from an Industry 4.0-compliant control device, but communication 'upwards' (northbound) is via the Industry 4.0-compliant asset administration shell using OPC UA. This architecture will also be adopted by the other fieldbus providers. Northbound, i.e. peer-to-peer control to controller, controller to MES or controller to cloud, OPC UA - and, if necessary with regard to hard real-time, TSN - will prevail.

A juicy detail in passing: in a similar ETG image published at the Hannover Messe 2015, the Ethercat Automation Protocol (EAP) was still shown northbound in addition to OPC UA. EAP has now given way to the sole northbound standard OPC UA. A change of strategy for the benefit of the user can happen that quickly. Copycats welcome!