Agreed, TJ. In 2011, Rockwell said that 60% of the automtion equipment they sold was Ethernet-based. Since then, I'm sure that figure has risen. The old complaints about lack of determinism are gone now, I believe.
Chuck, If you're like me, you can get in the habit of scanning an article, and missing some details as a result. The installed based numbers are really not that significant, in the end. The real news is the trends in new installations.
Chuck, I believe the first number reflects the expansion of Ethernet motion networking (to triple by 2016). The second numbers reflect Ethernet motion networking versus the total installed base of nodes, which means that there are alot of fieldbus nodes installed around the world so the market share number is dropping more slowly.
it seems like Ethernet networking will triple, but not at the expense of fieldbus technology. If Ethernet is rising so sharply, shouldn't fieldbus usage drop?
I think that was APresher's point - article is a little confusing. I believe the meaning is that for networked I/O, fieldbus protocols - DeviceNet, Profibus, ASI, etc - account for 75% of installations vs ethernet based protocols. This number is forecast to drop to 69%, so ethernet protocols are gaining ground.
A couple of things:
1. 2 - 3 years ago, I think ARC released a report indicating that Profinet and EthernetIP each had about 30% of all new ethernet based nodes being installed worldwide (I think EthernetIP had a slight lead). Has anyone seen more recent numbers? Both organizations claim momentum is on their side.
2. I don't think I would include EtherCAT in this category. I think EthernetIP, Profinet, Modbus TCP, etc, can tolerate other protocols running on their network. EtherCAT really can't. Well it can, but the performance difference.... I think it would be significantly worse than either Profinet or EthernetIP.
3. I'm amazed at how much share Modbus TCP continues to maintain. Is this just because pretty much anyone can do it with very little effort?
Important story, Al. I'm curious, though -- it seems like Ethernet networking will triple, but not at the expense of fieldbus technology. If Ethernet is rising so sharply, shouldn't fieldbus usage drop?