Death Knell of Fieldbus

Industrial Ethernet: The Death Knell of Fieldbus?
By Perry Sink
A year ago, a guy walked up to me at my spiffy
new Fieldbus display at a trade show and
proclaimed to me that the days of Fieldbuses
were numbered. He stood boldly in his $500
suit and proclaimed a new gospel to me, in the
tradition of the ancient prophets. The wrath of
the marketplace was fallen upon the heads of
Profibus, DeviceNet, Interbus and ControlNet,
he said. No longer would customers strive with
the wayward inclinations and collective
confusion of Fieldbuses. The writing was on the
wall: the king of TCP/IP and the pagan armies of
Ethernet were gathered outside the walls of the
city, he said. Our banquet feast was about to
come to an abrupt and bitter end.
Woe was me. My Dream of DeviceNet and the
Promised Land of Profibus would wither away
under the scorching sun of Ethernet. The
unfortunate part, I thought, was Gee, this
banquet feast weve been having hasnt even
been much of a banquet feast anyway. Its been
a more like an all-you-can-eat buffet at
Kentucky Fried Chicken. Most people havent
exactly made a killing in this Open Control
game yet.
The Case For Ethernet and a Single
Universal Standard
He did have a point.
Ethernet is
EVERYWHERE.
Matter of fact, millions
people use Ethernet when they rise up, when
they sit down, when they log in and when they
log out. And half of em dont even know
theyre using it. You can go to an office supply
store and throw Ethernet cards, software,
switches and routers in your shopping cart. The
guy at the checkout lane who takes your money
doesnt know the difference between a packet
and a pita pocket. But when you take the stuff
to your office and install it, it works anyway.
It logically follows that the same benefits could
be applied to automation systems. Then all this
chaos over protocols, proprietary networks, open
standards etc. would evaporate. Machine wiring
would be as open as the Internet itself. 20+
popular networks and hundreds of proprietary
homebrew protocols would crumble away, and
a single universal standard would emerge
triumphant. The next step, of course, would be

one-world government.
I think theres a
committee working on that, too.
But Then Theres Reality
Many manufacturers have already realized this,
and new Ethernet I/O products are appearing on
a monthly basis. And it almost seems that
everybodys got a different way of doing it. I
heard of one company who actually reverses the
position of two wires on the cable to make sure
the customer only uses their products on his
network.
The biggest issue is this: Ethernet itself is a
physical layer specification, which defines the
basic electrical characteristics of the network.
And TCP/IP is a way of packing data into
messages, arbitrating transmissions from
multiple sources, and ensuring complete reconstruction of a message at the other end of the
wire. It says nothing about what that packet of
data contains or what it means. It could be a
picture or a web page or a text document or an
encrypted message or 1 bit of data that makes
your front porch light turn on. And yes, it could
be I/O data from your conveyor system. But
theres no rule or standard as to how that data is
put on TCP/IP and taken off. Several methods
that are being used right now come to mind:
putting the Modbus RTU, ControlNet,
DeviceNet, Profibus, or HPIB (Hewlett Packard
Interface Bus) protocols on top of TCP/IP and
packing bits and bytes inside those protocols.
Plus others have rolled their own entirely.
It is possible to put a dozen different protocols
on one network and have them all talk to one
host at the same time, just as its possible to put
a dozen different brands of printers on the same
network, and print to any one you choose. The
problem is, you simultaneously need a dozen
drivers. Aint that fun? Im sure all of the
software vendors are looking forward to writing
them!
A Dozen Proprietary Protocols With The
Same Name?
So the current situation is this: Industrial

Ethernet, in reality, is a dozen different
proprietary designs, which are all called
Industrial Ethernet. So far, it doesnt really
solve the problems that a single universal
standard would solve. And it certainly doesnt
lessen the confusion.
Industrial Ethernet Association to the Rescue
To solve this problem, weve formed the
Industrial Ethernet Association, whose purpose
is to define and promote the use of Ethernet in
manufacturing and process equipment. The
following are among the issues which IEA will
address:
youve ever researched or installed a network of

any kind, you know its just like everything else
in engineering: there are tradeoffs at every turn.
Tradeoffs between speed and determinism, noise
immunity and distance, bit-level data vs. large
message capability, flexibility vs. ease of use,
etc. etc. We all know that no one approach is a
cure-all for everything. And no one networking
standard can address all needs in the
marketplace. There will always be multiple
standards, and when Industrial Ethernet is fully
defined, it will be another Fieldbus.
Finding Ethernets Proper Place among the
Fieldbuses
Ethernet has the following advantages:
A method of packing digital and analog I/O

data on TCP/IP with its associated
parameters and units
Standards for interfacing other complex
devices: drives, motion controllers, operator
interfaces, PLCs, barcode readers, etc.
Adopting an industrial-grade connector.
Everyone agrees that those plastic
telephone style connectors used in office
equipment wont survive the factory floor.
Adopting a deterministic architecture.
Identifying and clarifying the proper
interface between your machines Ethernet
and your companys intranet. Odds are your
I.S. Manager will start tossing bricks into
your cubicle if you mess with his network!
Ethernet has the following disadvantages:
Industrial Ethernet Association was founded by

Synergetic Micro Systems, Hirschmann,
Grayhill, HMS Fieldbus Systems, Hilscher
GmbH and Contemporary Controls. It was
launched in March 1999 has over a dozen
members now. The membership fee of $500 is
being waived for the first year. For more
information and a whitepaper on Industrial
Ethernet, visit www.IndustrialEthernet.com or
call (630) 434-1770.
Answering the Prophets of Doom: Ethernet is

Just Another Fieldbus
The prophets proclaim the death of Fieldbus.

Its a free country, so let them talk. But if
Its universally accepted.

The huge market for Ethernet products and
components makes them inexpensive,
widely available and thoroughly tested.
At 10 or 100 Mbit/sec, its fast.
Most engineers are already at least
somewhat familiar with it.
Its used in corporate LANs so its a natural
fit in the Sensor to Boardroom
connectivity scenario.
TCP/IP is also used on the Internet, so it
makes Web Enabled applications easier to
implement
Its not as practical to use it at the very

lowest level, i.e. embedding it in single
sensors, as it is with DeviceNet or ASi.
Its universal acceptance tempts users to try
to do too many things on one network and
presents security issues.
The connectors that are currently standard
do not meet the physical demands of
production equipment.
It is collision-based and not inherently
deterministic.
Its not as immune to noise as other
Fieldbuses.
It has a large amount of overhead and
doesnt use its bandwidth as efficiently as
other buses.
So when its time to design a control system,

Ethernet will go on the list and be examined on
the basis of its merits just like every other bus.
May the best man win! Certainly, different
networks will win in different applications.
fiber optics. But then the cost advantages go

out the window!
Prophetic Chrystal Ball
Its anybodys guess what will actually happen,
but heres what I expect to see in the next 2-4
years with Ethernet and the other Fieldbuses:
This Ethernet RJ45 telephone

connector is great for the office,
but would you put it on a robotic
welder?
Ethernet Perception vs. Reality:
In the general consumer market, perception is
reality. In the electronics world, perception
eventually becomes reality with the inevitable
progress of technology. But here are some
general observations about putting Industrial
Ethernet connectivity in factory floor products:
Its Dirt Cheap: When you embed

Ethernet into a product, youll find that its
not necessarily any less expensive to
implement than other buses. As a matter of
fact, its considerably more costly per node
than CAN based protocols like DeviceNet.
It requires a fair amount of horsepower.
Ruggedizing the physical connections will
also raise the cost.
Its Totally Open and Universal: This is the
perception, but as I discussed earlier, such is
not actually the case today. Eventually
confusion will give way to order and it will
be open and universal.
10 or 100 Mbits/sec is so blazing fast, speed
is no longer a problem: When you do actual
comparisons taking all factors into account,
this is often not true. At high speeds, noise
becomes a problem and the way out is to use
Ethernet will become a popular format for

rack and cell level networking and
above. It will not replace bit-level sensor
buses, though. It may actually stimulate
their growth by complementing them.
While the standards are being defined and a
cloud of confusion still surrounds Ethernet,
the existing Fieldbuses will continue to grow
and become less expensive.
Web Enabled controls products will
proliferate and software that runs on a web
browser will be the rage.
Some fringe networks will fall away (this
is already happening) and a half dozen
standards will be widely used, Ethernet
being one of them.
The cost of networking will continue to
drop, familiarity will become nearly
universal, and everyone will use networks to
connect everything.
So let the Ethernet armies march into our

banquet feast! And tell em all to bring a bottle
of champagne when they come. Therell be
enough Food, Fun and Fieldbus for everyone.
Available on request: The 24 page white
paper, Industrial Ethernet and 8 Popular
Fieldbuses. Email [email protected] to
receive a copy.
Perry Sink is Training Manager at
Synergetic Micro Systems in Chicago
(www.synergetic.com / (800)600-0598, a
manufacturer of industrial networking
hardware, software and tools including
Ethernet.
Originally published in Manufacturing Automation Magazine,
Canada, April 1999 2001 Synergetic Micro Systems, Inc.

Death Knell of Fieldbus

Uploaded by

Copyright:

Available Formats

Death Knell of Fieldbus

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Death Knell of Fieldbus

Uploaded by

Copyright:

Available Formats

Industrial Ethernet: The Death Knell of Fieldbus?

triumphant. The next step, of course, would be

So the current situation is this: Industrial

youve ever researched or installed a network of

A method of packing digital and analog I/O

Ethernet has the following disadvantages:

Industrial Ethernet Association was founded by

Answering the Prophets of Doom: Ethernet is

The prophets proclaim the death of Fieldbus.

Its universally accepted.

Its not as practical to use it at the very

So when its time to design a control system,

fiber optics. But then the cost advantages go

This Ethernet RJ45 telephone

Its Dirt Cheap: When you embed

Ethernet will become a popular format for

So let the Ethernet armies march into our

You might also like