Sorry, but I don't see this as solving "industrial Ethernet" problems. Perhaps "industrial networks." Many of the interoperability and security questions are either solved, or more accurately "continuously being addressed," by the IP layers.
One additional feature that could potentially be folded into an "industrial Ethernet" would be to introduce a synchronous or isochronous mode, at layer 2. But this attempt is what was the downfall of a lot of other layer 2 protocols that have been left in the dust. So instead, we have solutions like RTP/RTCP, which address that problem at layer 3.
I think there's a persistent underlying misconception of what Ethernet actually is. In a "big picture" kind of way.
Okay, so that one aspect could really be an "industrial Ethernet" new direction. Parenthetically, I've seen attempts at this literally for decades. Almost from the time Ethernet was first standardized. Even if these efforts were not IEEE. They were a staple for academic theses, back in the '80s and '90s. It was also attepted in FDDI, where synchronous mode was never mush used or understood.
But glomming on a lot of non-Ethernet topics into this "industrial Ethernet" forum is very confusing to me. Let's take something most people know intrinsically. The plumbing and HVAC systems in their homes. So, this new forum sounds to me like it's going to address industrial HVAC installations for the future, including topics on the design of bathroom faucets, toilets, and pressure regulators.
Of the five goals listed, security jumps out as Numero Uno. We just can't have people hacking into corporations, banks, water systems, electric grids, and more. Adding more layers of security simply isn't working. The problem is that the Internet, because of its current structure, can only play defense. What is needed -- and I have no idea if this group would address is -- is an offensive capability that can track down intruders. I often read that's impossible, but anyone who's lived through the last half century knows that nothing is impossible.
Broader question: I get the need for an industrial internet. But why is it only an "industrial" Internet? Why not improve all those things on THE Internet?
@Tom: As I am coming to underdstand it the thing we looosely call Interneyt of Things is more broadly known in the research community as the field of Cyber Physical Systems. In this field CPS needs are seen as distinct from those of the consmer Internet. Sunder suggested CPS is a superset of the commercial sector GE refers to as the Industrial Internet, which is itself quite broad spanning automotive, mil systems, the smart grid and etc--but not the IT field you are headed to at InfoWeek ;-)
Industrial Internet could revolutionise the industry only if there is real good analytical software to analyse the data and really help the machines to perform efficiently and also help the actual users.
Great explanation, Rick. Thanks! Only problem is: now my head hurts. Are we moving to an era of tiered Internet service in which some will be faster and more secure than the Internet most people use every day? I'm not sure what that means -- and maybe the consortium doesn't yet, either. EG: Does that involve verifiable log-ons that are not available to you and me?
I can recall a news conference at Intel back in, oh, '95? '96? It was in the start of the Information Superhighway hype, and tales were spun about how a consortium of corporations and the government was going to make the Internet "super." Al Gore was piped in by video. Very impressive. And Gore was later inducted into the Internet Hall of Fame for his efforts. So I bring a bit of baggage to this latest announcement.
Did the earlier effort really make a difference? Did the Internet change business or education or healthcare because of that? Or just because it would have done that anyway? Will the new consortium really make a difference? Or is the Internet of industry and the Internet of things going to happen anyway?
"Why not improve the current Internet itself to make it industrial grade?"
This is happening all the time, though. The RFCs published by the IETF seem to be exponentially increasing, and these include security topics. That's part of my questioning what this new initiative is all about. What would be different for an "industrial Internet," having to do with Internet Protocols, that isn't already being done in the numerous IETF working groups?
Not much that I've seen listed as goals, at any rate.
I am lost here...what is this Industrial Internet about? ...is this a separate physical Internet infrastructure with higher level of security? or a sub-set of exisiting Internet network available to some users? Kris
You're not alone. When I read the title, I thought they are trying to recreate a private Internet for the industrial market. What does industrial market mean anyway? As I read the article, I'm lost more. I thought all those problems are being addressed. To dig deeper, I went to the GE white paper. The white paper seems to serve a marketing campaign to me. I guess I definitely need some elaborations and insights.
I can hardly disagree with you. What's the difference between Internet of Things and Industrial Internet? What the consortium should address are in my opinion (1) Security of internet of internet of things (2) software reliability in aspect of real-time in this era of cyberphysical systems of mobile devices, home devices, vehicles and industrial systems whilst inter-connecting/inter-operating. Although, if for the creation of new and distinct standards for industrial, i think a worthy one. Nevetheless, security should top framework agenda.
I have been involved in industry organizations in the past that did not have clear direction and they rarely went anywhere. One big flashing red light for me is the desire to "...gain a competitive advantage" through this initiative. Isn't this the kind of thing that we have seen China trying (and generally failing at) in telecom and other domains?
There are problems to be addressed. Security is an issue, as Tom and others have pointed out, but that is not just an industrial issue. It might be viscerally tempting to implement counterattack as he suggests, for example, but given the nature of many attacks that would be a bad idea. If actor A attacks site B through site C, then the counterattack would hit C rather than A. Attribution is a real problem in this space.
The real trick is to implement and of this without breaking compatibility. A secure industrial internet that does not coexist with The Internet would have real limitations in terms of capability. You can do pretty much the same thing with an appropriate air gap.
The Internet is fine. It's the Internet of Things that needs fixing.
IoT lacks what the Internet has-- a unified set of standards like IP, HTTP, HTML and etc so lots of people can do lots of great stuff. IoT is a Babel of protocols and proprietary implementations. It's a mishmash in which no one node knows how to talk to another one.
You can view this Web page on your iPhone, Android, Windows PC, Mac or Linux box because the Internet is GREAT. It has a common set of standards. It does NOT need fixing.
The Internet of Things, however...it needs major work.
Blog Make a Frequency Plan Tom Burke 17 comments When designing a printed circuit board, you should develop a frequency plan, something that can be easily overlooked. A frequency plan should be one of your first steps ...