March 20, 2009

No Need to Lock Them in a Room, Mr. Secretary—Use IP


A lot has been said and written this month about the need for speedy adoption of interoperability standards for smart grid, but the simple fact is: Agreeing on interoperability standards doesn’t need to be this hard. Incumbent manufacturers of some niche smart grid technologies based on proprietary protocols and those companies that have built siloed systems using them have every incentive to drag out the process of identifying and agreeing on interoperability standards for smart grid. But, the answer is so obvious; it seems hard for some people to grasp: Use Internet Protocol (IP).

At the center of the Internet boom in the 1990s were not only the price declines in computing and telephony, but also the transition from the use of proprietary protocols and standards to the adoption of open protocols and standards – namely IP – that allowed communications devices and systems to interoperate, spurring innovation and promoting competition. IP was an important factor in enabling the market disruption that drove Internet use – all to the benefit of consumers. Without IP, we wouldn’t have YouTube, or Facebook, or Internet banking.

Who wants to wait in line at the bank anymore?...

Today’s public discussion about interoperability and smart grid reminds me of that time. Smart grid is still relatively new. To me, smart grid is about adding intelligence into the nation’s electricity delivery system, much like how telecom service provider networks went from being voice-only networks to providing all kinds of new value-added services that go along with the voice service into your home or business. With its open architecture, IP networks can scale, manage, interoperate and secure communications in ways that no network built on proprietary technologies can. And we’re not talking about running smart grid over “the open Internet.” Indeed, many IP networks are private networks, with strong security that protects not only the data, voice and video as it move through the network, but also provides the physical security necessary to protect the grid’s critical infrastructure.

Section 1305 of the Energy Independence and Security Act of 2007 (EISA, P.L. 110-140) designates the National Institute of Standards and Technology (NIST) as having, “…primary responsibility to coordinate the development of a framework that includes protocols and model standards for information management to achieve interoperability of smart grid devices and systems.  Such protocols and standards shall further align policy, business, and technology approaches in a manner that would enable all electric resources, including demand-side resources, to contribute to an efficient, reliable electricity network.” Sound familiar? NIST began its smart grid interoperability coordination work soon after EISA was passed in December 2007. While progress has been slow, it’s been steady, and we expect it will accelerate given recent comments by Secretary Chu, which have been reported widely in the press.

Tempting as it may be to lock everyone in a room until they come to agreement on smart grid interoperability standards, it really doesn’t need to come to that. Cisco can help make this process short and sweet: Use IP.

Jennifer Sanford Posted by Jennifer Sanford at 01:33PM PST

Permalink, Comments (10), Trackbacks (0)

Tags: internet protocol interoperability ip smart grid standards

10 Comments

Akiba Mar 22, 2009

Hmmm…your writing is good, but the post is a bit underinformed. IP only specifies a communications protocol and doesn’t specify how devices interact with each other. The reason you can communicate via TCP/IP from a server from your PC is because standards exist for HTTP, HTML, CSS, and others and both the servers and the browsers agree to conform to them.

In order to have device interoperability with TCP/IP, you would need to establish common behavior for all the devices such as what the frame formats would be above the TCP layers, how the devices respond to requests, what types of requests the devices can accept, etc. On top of that, you would need a way to test and enforce interoperability between all the device manufacturers which involve drawing up testing guidelines and certifying 3rd party test houses.

There is currently a movement to use TCP/IP over wireless sensor networks using 802.15.4 and an IPv6 header compression scheme. It’s called 6LoWPAN and it’s currently and IETF draft. However they are struggling with these issues as well as how to accomplish routing, device discovery, and service discovery.

In my blog, I argue that the Zigbee Alliance and the 6LoWPAN camp should work together and combine the benefits of TCP/IP ubiquity with Zigbee’s work at device interoperability including creating a Smart Energy Device Profile and lining up manufacturer support.

If Cisco wants to take a leading role in this, why don’t you mediate between the two camps and work with them to implement your vision of using IP for the smart grid.

Otherwise, your comments don’t really help the situation and just repeat what’s already known in the industry.

the cisco kid Mar 25, 2009

I want work on these kinds of issues to be my life’s work. I think this means I have to go to law school. The article and comment are informative, I will be sprawling the links provided.

thank you.

Jennifer Sanford Mar 25, 2009

Thanks for your insightful comments. Oftentimes, repetition of the facts is a good strategy when the debate veers off topic. While I purposefully focused on IP as the preferred inter-communications protocol for smart grid, your comment actually refocused the IP discussion on TCP/IP in Home Area Networks. Many seem to equate smart grid with smart meters, while at Cisco, we view smart grid as a communications platform that spans the entire grid ecosystem.

As you know, TCP/IP is but one protocol in a family of IP protocols (including UDP, ARP, DHCP, SNMP - among many others). In raising TCP/IP, you graciously made my point for me: One of the strengths of IP is that it is a family of protocols that coexist and interact to bring about communications between devices and applications. While IP does not directly address device interoperability, it does address inter-communication. Indeed, as you say, the business logic is in the protocols carried over IP, and getting people to agree on the business logic entails a whole lot of work and time… So, it would appear to me that in fact, our arguments seem to complement one another.

With respect to smart grid interoperability, it’s certainly debatable whether regulators are more concerned with substation interoperability or interoperability in the HAN. I won’t venture an answer to that, but it occurs to me that no one is talking about using Zigbee anywhere other than the home. In the substation context, disparate protocols are presently used by utilities in SCADA networks, such as DNP 3.0, Conitel 2020, SES91 & 92, Modbus and a host of others.  In SCADA, as can be found in other industries - such as process control, computer networking, physical security, video surveillance, building controls, etc. - non-IP protocols begin a migration towards convergence by first designing an IP gateway to tunnel the non-IP protocol in such a manner as to enable native IP devices to mimic the combination of the non-IP device + the IP gateway. Once the gateway is designed, the convergence process continues through encapsulation within IP, then bridging to IP and finally becoming native IP. We believe that the next generation of devices will favor the 1-step (IP) architecture over the 2-step (non-IP device + IP gateway) architecture.

Why is this market so fragmented? First, the plethora of proprietary or semi-closed protocol stacks make the communication between devices only possible by means of protocol translation gateways that have made it unavoidably difficult to manage and operate, and have added lots of unneeded complexity. By the using IP, this major issue is solved first, making the communication between devices significantly simpler, less expensive and more robust. It is also worth mentioning that this makes the use of various media possible thanks to the layered architecture of the TCP/IP protocol suite. We surely agree on the fact that applications running over TCP-UDP/IP need to be standardized or simply adapted to the use of IP.

The point about IP versus the tower of Babel of non-IP protocols is simply that rather than trying to perpetuate the silos, let’s promote the migration that is already beginning with products, such as the OrionLX (http://www.novatechweb.com/orion.cfm), and move the industry to IP. 

In consulting with one of our Distinguished Engineers, JP Vasseur, on your statement about the movement towards the use of TCP/IP over wireless sensor networks using 802.15.4 and an IPv6 header compression scheme, JP reports: "Header compression techniques of IPv6 packets have been specified at the IETF (not a draft, but an RFC4944), and other schemes are in the works. As far as routing is concerned, major progress is being made to specify a routing protocol for these constrained networks (the working group is ROLL: http://www.ietf.org/html.charters/roll-charter.html). Please note that the routing solution is being defined regardless of the media, thanks to layer 3 routing, and is not tied to 802.15.4 in any way. With regard to device and service discovery, there are many IP-based protocols that could be reused here."

With respect to your suggestion about Cisco taking a leadership role in mediating between the Zigbee Alliance and the 6LoWPAN camp, we certainly have been advocating the use of some of the Zigbee profiles, such as the Smart Energy Profile over an IP stack. 

Perhaps I am repeating the message about IP and smart grid, but I would argue that it’s an important message for industry, regulators and policymakers to hear again and again. Plus, it beats getting locked in a room by Secretary Chu...

Thanks to the strong teamwork and collaboration we enjoy at Cisco, I’d like to acknowledge and thank the following colleagues for their contributions to this posting: JP Vassuer, Cisco Distinguished Engineer; Rick Geiger, Cisco Director of Business Development for Enterprise Utility Solutions, and; Burt Welsh, Cisco Program Manager in the Corporate Development Technology Group.

Akiba Apr 2, 2009

Thanks for the detailed reply and you’re indeed correct. I was focusing on the consumer end in the home, and missing the whole discussion about the backend.

After the whole smart grid hacker debacle, I was informed anonymously that the vulnerability was from the SCADA on the backhaul which allowed the worm to propagate. That whole issue forced me to take a look at the grid as a whole, whereas I’ve mostly been focused on the meter side which is where all the Zigbee action is.

In regards to IP over Zigbee, I also hope that it becomes a reality because if both are paired, then it can take wireless sensor networks to a new level in the consumer marketplace. It would also put an end to the fragmentation going on so the industry can focus on creating actual products. Right now, everyone is just sitting on the fence which is killing alot of the startups.

As for the backhaul that you were referring to, I’m not really knowledgeable about that end and if that was the main focus of your post, please disregard my comments. I seemed to have been in a pissy mood that day…

Joe Apr 13, 2009

It looks like Ambient corp is already using IP as the standard. 

http://www.ambientcorp.com/smartgrid.html

“The Ambient Smart Grid™ communications solution is a modular network overlaid on the medium-voltage and low-voltage segments of the power distribution system allowing real-time insight into the operations of the electrical distribution grid while supporting any IP-based application. High-speed backhaul connections (which go to the Internet or to private networks) connect the Ambient Smart Grid™ network at any point along the medium voltage circuit allowing for IP data traffic to be carried, via a choice of multiple technologies including BPL, Wi-Fi, cellular, and/or low bit rate power line carrier (PLC)”

Jennifer Sanford Apr 30, 2009

Yes, Joe. Indeeed, several companies are also promoting IP; I would point you to the IPSO Alliance (http://www.ipso-alliance.org), which Cisco co-founded with Sun.

Houston Technology Lawyer Oct 5, 2009

It will certainly be interesting to see how smart grid systems evolve. As Akiba mentioned above regarding set standards can be a real obstacle. Though an independent organization develops and evolves web standards, not all browsers are fully compliant which can be a pain when developing websites which don’t look the same across all browsers. Getting everyone to agree to conform to set standards can make it easier for progress to occur.

Web developers Oct 15, 2009

That was inspiring,

IP is the reason why we have facebook and youtube!!  what a discovery!

Keep up the good work

Giovanni Oct 16, 2009

Jennifer,Thanks to the strong teamwork and collaboration we enjoy at Cisco, I’d like to acknowledge and thank the following colleagues for their contributions to this posting: JP Vassuer, Cisco Distinguished Engineer; Rick Geiger, Cisco Director of Business Development for Enterprise Utility Solutions, and; Burt Welsh, Cisco Program Manager in the Corporate Development Technology Group.

Colon Cleanse Oct 19, 2009

With its open architecture, IP networks can scale, manage, interoperate and secure communications in ways that no network built on proprietary technologies can.

Post a comment

Join the conversation!

We encourage your comments, questions and suggestions. All comments are moderated and will appear as soon as they are approved by the moderator.

Please increase the validity of your comment by providing a valid first and last name. Spam, off-topic or offensive comments will not be posted.

Name:
Email:
URL:

Comments:

Notify me of follow-up comments?

Submit the word you see below:


Post a trackback

Ping this URL to post a trackback:
http://blogs.cisco.com/trackback/7126/eY9WCi82/

More blog posts

Previous post:
Latin America Is Going Fiscal At Full Speed

Next post:
Chimpanzees smarter than undergraduates and college professors?

Recent posts:
November 2009 Archive