../blog/of/rADHIoflyer

1 Introduction
Monitoring and controlling sewage plants, well fields, pipelines, speed displays on motorways or train signals can now be integrated in one Industrial Ethernet. To benefit from the advantages, a choice has to be made between competing technologies and solutions. Instead of backing a single solution at this early stage, it’s more advisable to extend and harmonise existing infrastructure gradually.

When setting up, it gets complicated when the local control units have to communicate with planning systems higher in the hierarchy. For example the service technician goes to sewage plants, without knowing what faults to expect. Or a decline in pressure is reported in a gas pipeline and the fault cannot precisely be localised. Once again the creativity of engineers is called for. There are LAN-to-field bus-solutions, modem connections and manual data transmissions with lists, faxes or copying and pasting between different applications. Particularly annoying are the widespread batch processes for transmitting real-time data.

2 Departing from Ethernet and migrating to Industrial Ethernet
As separate data in the administrative processes and machinery to be monitored turn attempts towards real-time processing on both sides into a farce, the idea of Industrial Ethernet was born. A network for everything, allowing rapid transfer of information with a single protocol and suitable for heavyduty use under extreme ambient conditions. As Ethernet is available in different types and speeds, for copper and optical fibres, for local and remote networks, it’s an obvious choice. But the path from Ethernet to Industrial Ethernet is longer than many people think. The demands on an Industrial Ethernet Network are a lot tougher than in an office. Just ask a systems administrator if rain or snow, or vibrations from heavy goods traffic are acceptable in the server room. And then eight to ten hours a day. But the network cables don’t mind if the machine next to them needs a high-voltage current, do they? The second problem is the way Ethernet works. All the nodes share one cable. If two devices transmit at the same time, there’s a collision that briefly delays data transmission. It’s not important for an email if the network card is forced to break for 50 or 100 milliseconds, it will try again later. But for realtime applications this simply isn’t acceptable. For field bus solutions with SCADA-information (Supervisory Control and Data Acquisition) a latent period of five to ten milliseconds is deemed acceptable. Small network segments that just a few systems are connected to, only afford some protection, for example against an e-mail that is blocking the system because of its PowerPoint presentation attachment. Despite all the hurdles, the majority of problems with Industrial Ethernet have now been solved. But there isn’t just one single Industrial Ethernet. This is more of a generic term that describes a collection of the most varied technologies and solutions. Ethernet-Frames and TCP/IP are the elements that unite them. But 14 different protocols alone are competing to become the standard for real-time Ethernet applications. Some solve general problems, others simply concentrate on connecting field bus systems. Industrial Ethernet can be divided into two segments. On the one hand Industrial Ethernet is restricted to factories, so when infrastructure is set up an optical fibre can quickly be fitted. On the other hand, Industrial Ethernet is used on narrow, long pieces of land. When setting it up, the technicians have to use the cable that’s already there – regardless of whether it’s optical fibre and copper cables.

At the same time there is already a wide range of hubs, switches and media converters, adapted to industrial conditions in both the segments. On the one hand there must be more resistance to ambient conditions such as dust, spray from water and fluctuations in temperature. On the other hand the emphasis is placed on reliability and protection from shaking, fluctuations in voltage, strong electrical fields etc. The connections are also different to an office network. Direct current (usually 24V DC) is often applied and 35 mm DIN top hat rails (support rails) are used to fix it.

As the network connection in the local and energy supplier environment often occurs from sensor to sensor in a row, Industrial Ethernet networks are often set up in a ring, that can consist of a transport system mix with SDH network structure and its own transport system. For example linear network structures are used particularly in railway signalling, monitoring on motorways and pipelines. By intelligently using the structure and with appropriate alarm systems, a disruption can be restricted to a maximum of one switch and appropriate countermeasures can be taken. If a line fails, the network can carry on functioning. The Rapid Spanning Tree or one of the previously mentioned, non-standardised protocols that permit a quicker response time, will disconnect the ring network.

The change from different communication structures to Industrial Ethernet provides a range of benefits. The most important is the use of a single standard to connect all types of network. The boundaries between LAN, WAN and technical network disappear, devices from different manufacturers can be connected with one another. At the same time, technology in local and energy supplier networks can be used that have already been tried and tested in other areas. For example remote maintenance via integrated web servers over HTTP. Instead of many different applications to control different regulators, CCTV cameras and sensors, a single Web browser could soon suffice. Service technicians could dial into a virtual private network (VPM) and have the status of a gas pressure sensor displayed, in order to bring the right spare parts instead of ordering them after carrying out maintenance. Just as in automotive technology, it would be possible to equip control units with flash memory in order to collate measurement data over a longer period of time in the device and then send it via FTP or SNMP for analysis. Ethernet provides the bandwidth required to deal with larger amounts of data. But this is just a dream for today’s users.

In order for Industrial Ethernet to unfold its full potential, thinking way outside the box at the design stage – beyond a single sewage plant or a railway station – is vital. The aim is to exceed all artificial boundaries and interconnect the whole company with a homogenous network. A small set of flexible protocols can then open up new applications. This can happen in several steps. It’s better to exchange everything than to have exchange between everything.

3 Planning the introduction of an Industrial Ethernet
Planning the introduction of an Industrial Ethernet starts when information is collected or monitored, but shouldn’t end there. What happens for example with connecting remote networks? Nowadays many companies rely on connection via modem to the public telephone network. But telecommunications companies already supply a variety of services for coupling IP-networks without a media crush that can be implemented without dedicated leased lines. Synchronous DSL (SDSL) with a guaranteed velocity of two mbps per channel is already offered in many conurbations. But the networks might have to be extended on the company’s own site too. Many companies favour optical fibres, because they’re more robust and are, in contrast to copper cables, not dangerous in thunderstorms. They seem to be a good choice in railway or gas pressure stations because the optical fibres are resistant to electrical fields from other machines. But laying new cables always means high investment. Additionally, in mixed networks consisting of copper and optical fibres, media transmission must be ensured. Because the market is currently so fluid and solutions compete with one another in many areas, it’s better to keep your options open. It’s usually less expensive to use the existing infrastructure and specifically extend it. Because the aim of Industrial Ethernet is not to introduce a completely new network, but to extend the existing one sensibly, to strengthen company procedures. The nodes are particularly important that guarantee the connection of the networks. In this case flexibility, quality and long-term reliability is worth it. Because nothing is more expensive than building a new infrastructure around the connecting points. It’s much better to choose systems solutions that also

support new solutions. As soon as for example a continuous IP-network exists, it’s worth considering whether telephone traffic could be handled between your own locations (Voice over IP.) Or perhaps it’s a good idea to transmit video data via CCTV cameras to joint safety headquarters? But then copper cables could reach their limits. Luckily service providers supply direct connection to SDH-optical fibre networks (synchronous digital hierarchy.) The question remains whether your connection technology will also handle the change from copper to optical fibre to third transmission systems. Whatever the future might bring, there’s no way of avoiding Industrial Ethernet. Because of the dynamic market it’s still difficult to foresee all the developments. The motto has to be: Industrial Ethernet – definitely, but only based on flexible connection nodes that won’t place restrictions on the user in
future either.