I've asked and been asked that question myself a number of times, and while replacing purpose-developed field instrument networks with Ethernet has a certain elegance in that it would consummate Ethernet's conquest of the industrial networking realm, the obstacles as I understand them are several.
Historically, instruments have resisted Ethernet's infiltration because of the need to provide power over the same set of wires as the signal (loop-powered), a standard long-required by the process industries for a digital alternative to 4-20mA.
More recently, the emergence of Power-over-Ethernet standards would appear to have solved this problem, but Ethernet includes a lot of network overhead that is overkill for a pressure transmitter--and that overhead consumes bandwidth and more importantly power. So, if you think of multiple transmitters on a fieldbus segment -- more power consumption limits the number of devices you can put on a segment, especially if you need to ensure intrinsic safety. One strike against them.
Further, Ethernet chipsets in field instruments would require a lot more than reworking of the instruments themselves -- the entire process instrumentation infrastructure would have to be re-engineered, and to little ultimate advantaqe in terms of cost or performance.
Ultimately, it appears there will be a place for purpose-built networks as a complement to Ethernet for quite some time to come. How long is anybody's guess, but hey--there's still a lot of pneumatic instrumentation out there!
I would be curious what other perspectives are out there.
Perhaps a matter of time -- but still quite a while
Scott -
I've asked and been asked that question myself a number of times, and while replacing purpose-developed field instrument networks with Ethernet has a certain elegance in that it would consummate Ethernet's conquest of the industrial networking realm, the obstacles as I understand them are several.
Historically, instruments have resisted Ethernet's infiltration because of the need to provide power over the same set of wires as the signal (loop-powered), a standard long-required by the process industries for a digital alternative to 4-20mA.
More recently, the emergence of Power-over-Ethernet standards would appear to have solved this problem, but Ethernet includes a lot of network overhead that is overkill for a pressure transmitter--and that overhead consumes bandwidth and more importantly power. So, if you think of multiple transmitters on a fieldbus segment -- more power consumption limits the number of devices you can put on a segment, especially if you need to ensure intrinsic safety. One strike against them.
Further, Ethernet chipsets in field instruments would require a lot more than reworking of the instruments themselves -- the entire process instrumentation infrastructure would have to be re-engineered, and to little ultimate advantaqe in terms of cost or performance.
Ultimately, it appears there will be a place for purpose-built networks as a complement to Ethernet for quite some time to come. How long is anybody's guess, but hey--there's still a lot of pneumatic instrumentation out there!
I would be curious what other perspectives are out there.
Keith Larson