Image Processing Reference
In-Depth Information
Given the multitude of approaches, despite the early hope that Ethernet could be the basis for a
unique industrial communication solution, the situation has not too much changed compared to
the heterogeneity of the traditional fieldbus systems. Interoperability between different Industrial
Ethernet solutions is not possible in a direct way and must be established on a higher level by means of
profiles (which is actually done to allow for cooperation between “old” ieldbusses and “new” Ethernet
installations) or middleware layers like OPC. he only conceivable improvement compared to field-
bus technology is that even with all proprietary modifications, Ethernet and to a large extent also the
IP suite are being recognized as technological basis for the new generation of industrial communi-
cation systems. And with a view on vertical integration, the main benefit of Industrial Ethernet is
that all approaches allow for a standard TCP/UDP/IP communication channel in parallel to field-
bus communication. Even the RTE solutions (like PROFINET, Ethernet Powerlink, EtherCAT, etc.)
have such a conventional channel for configuration purposes. his is related to the IP-over-fieldbus
tunneling concepts discussed earlier, but unlike the classical fieldbus systems, where such IP tunnels
were introduced long after the fieldbus development and thus often had to cope with performance
problems, they are now an integral part of the system concept right from the beginning on.
The separation of real-time and non-real-time traffic is accomplished on Ethernet MAC level with
prioritization or TDMA schemes together with appropriate bandwidth allocation strategies. In such
a parallel two-stack model, IP channels are no longer stepchildren of industrial communication,
but offer sufficient performance to be used for regular data transfer. While this enables in princi-
ple the coexistence of automation and nonautomation applications on Industrial Ethernet segments,
the mixing of automation and office is not advisable for performance, but more important for secu-
rity reasons. he value of this standard IP channel is rather to be seen in a simple direct access path to
the field devices. herefore, the currently favored solutions for configuration tools (i.e., XML, SOAP,
and more generally Web technology) can be used consistently. his again does not mean that Indus-
trial Ethernet solutions are interoperable or use the same configuration tools, but at least the basic
principles are the same.
Actually, all this could have already been done with traditional fieldbus systems as well, and it
certainly would have been done had especially the achievements of the Internet and the WWW been
available in the early s. So, what we see today with the rapid evolution of Ethernet in automation
can in fact be regarded as a second wave of fieldbus development which takes into account all the
technological achievements of the last decade and exploits them for the field level, thereby making
particularly vertical integration significantly easier.
20.8 Aspects for Future Evolution
Fieldbus systems have come a long way from the very first attempts of industrial networking to
contemporary highly specialized automation networks. What is currently at hand—even after the
selection process during the last decade—nearly fully covers the complete spectrum of possible appli-
cations. Nevertheless, there is enough evolution potential left [,,]. On the technological side,
the communication medium itself allows further innovations. he currently most promising research
field for technological evolution is the wireless domain. he benefits are obvious: no failure-prone and
costly cabling and high flexibility, even mobility. he problems on the other hand are also obvious:
very peculiar properties of the wireless communication channel must be dealt with, such as atten-
uation, fading, multipath reception, temporarily hidden nodes, the simple access for intruders, and
many more []. Wireless communication options do exist today for several fieldbusses []. Up to
now, they have been used just to replace the conventional data cable. A really efficient use of wire-
less communication, however, would necessitate an entire redefinition of at least the lower fieldbus
protocol layers. Evaluation of currently available wireless technologies from the computer world
with respect to their applicability in automation is a first step in this direction. Ultimately we can
