Blog Post

A Dozen Things the Smart Grid Can Learn from the Internet

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Balaji Natarajan is a senior IT strategist for Capgemini, focusing on smart grid, mobile computing, and unified communications.

History is a great teacher. The Internet took decades to evolve and take shape. Now upon us is a more critical challenge: how to build and sustain a smarter power grid. The tech community is already starting to initiate significant efforts to address this challenge, but here are some thoughts for how we can learn from the history of how the Internet was built, which will help us move into the fast-lane to tackle this massive task and monumental opportunity.

1. Scalable, Service-Driven Smart Grid Architecture: The smart grid needs to develop architecture that can scale up quickly, driven by the same kind of on-demand service-based models by which bandwidth is sold on the Internet. Like the Internet, the smart grid represents a combination of systems at an unprecedented scale: supporting millions of users demanding access to services that must be responsive, robust and always available. The number of grid communication sessions and grid access requests by interacting components (producers, consumers and gateways) translates into a very high number of network requests, placing an enormous demand on resources.

It is well-documented that power systems experience huge variations in service load, with bursts coinciding with the times that the service has the most value. In the early days of the Internet, the network relied upon the scaling capabilities of the operating system UNIX. Later, several systems capable of handling high-volume transactions were developed and coined as “Internet Servers.” Similar capabilities need to be provisioned for servers of the smart grid.

2. Defined Communication Protocols: The smart grid needs to standardize around a set of protocols and, importantly, define a key protocol for smart-grid communications. It could even take a cue from, and if necessary re-architect, the Open Systems Interconnection Reference Model to adapt to the “Internet of Things.” In the case of the Internet, the Internet Protocol Suite (also referred to as TCP/IP) enabled a common inter-network protocol and, instead of the network being responsible for reliability (as in the ARPANET), the hosts became responsible. With the role of the network reduced to a bare minimum, it became possible to connect almost any network, with any characteristics, together. A similar protocol design will be essential for the smart grid, given the immediate need to connect diverse component types.

3. Security & Encryption: The smart grid community needs to publish detailed specifications for different levels of security and encryption standards. For secure communication purposes, the Internet leveraged data encryption standards, including variants of Digital Encryption Standard (DES) as endorsed by the National Institute of Science & Technology (NIST). To ensure secure user authentication and data integrity, techniques like digital signatures and the network authentication protocol Kerberos were also created. Some of these standards are openly available whereas certain encryption tools are patented and require licensing.

Similarly, smart grid security needs to be thoroughly investigated to enable a multi-tiered security model for the grid. Once this is done, startups should be encouraged to build innovative tools that adhere to these standards. It’s important to note that the smart grid’s cyber-security layer may need to be more regulated (by federal policies) than the Internet’s has been, given the potential direct impact on national security systems. (See more details in No. 9.)

4. Management & Communication Tools for Energy: Simple HTML pages publicized the concept of the Internet to the common user back in the mid-1990s. A tool like this that offers a rich user experience can help in connecting the customer to the concept of the smart grid.

5. Open APIs: Companies building the next generation of the smart grid need to regularly publish APIs through working-groups and to meet specification requests. As they have done for the Internet, open APIs would ensure ease of integration of the grid with different endpoints, like smart meters, advanced metering infrastructure (AMI), programmable logic controllers (PLC), wireless mesh networks and portal gateways.

6. Find the Killer App: The smart grid needs to clearly evangelize its first killer app that could appeal across all grid user communities, including customers, utilities, the federal government, local governments and business sectors. Email did the trick for the Internet. So what will it be for the smart grid? Demand response? Energy management?

7. Connect the Grid to Consumers: Extend the functionality of the smart grid into a variety of always-on lifestyle interfaces, including meters, panels, garages, vehicles, recharging stations and mobile devices. The Internet really took off within universities (and then everywhere else) because of the concept of the “browser.” From that point on, Internet access wasn’t tied to an IBM supercomputer or a Windows proprietary desktop. With that shift, the tantalizing possibilities of open access invited strong investment, and we continue to reap more benefits of such a model more than two decades later.

8. Make a Task Force: The smart grid’s standardized framework needs to be driven by an engineering task force, similar to the Internet Engineering Task Force. Like the IETF, the smart grid task force (SGTF) would produce high quality, relevant technical and engineering documents that influence the way people design, use and manage the smart grid in such a way as to make the smart grid work better. These documents would include protocol standards, best current practices, and other informational documents. When the informational tools are dispersed, let application innovation (driven by businesses) drive the growth.

9. Allow for Secure Grid Layer: OK, this one is a little specific, but still quite relevant given the Internet’s roots with the Department of Defense. Because power grid security is so important, the smart grid needs to build policies and procedures that allocate a limited percent of a grid layer to be fully regulated and completely off limits for business access. This layer can be used for Pentagon/DoD grade security for administering, monitoring and controlling the resources and functions of the grid. It would basically be a virtual nerve center with selective access. While the Internet doesn’t have a nerve center per se, certain secure transmissions leverage the same distributed collaboration-architectures, and are fully controlled for classified access only — e.g., command and control communications.

10. Recovery in a Disaster: In the event of an emergency or disaster, the smart grid needs to have a fault-tolerant channel and ensure that all resources are set up for collaboration to deliver an automated, self-healing recovery operations process. This channel should not disrupt any other basic channels of communications or utility services. In the early days of the Internet, the network was successfully able to keep basic telephony services running during network disasters.

11. Government Labs: Over the years, as the smart grid matures in the open market, it is important for a smart grid governing body to still own and maintain state-of-the art lab environments to focus on the technologies involved. This ensures that the smart grid will still be current, open and standardized and not influenced too greatly by specific vendor interests. In the case of the Internet, the Department of Defense created ARPANET, which was helpful in the development of early application-services like email, FTP and protocols like NCP and TCP/IP.

12. Demonstrate the Value: Finally, but probably most importantly, the smart grid needs to develop and clearly communicate and market its value proposition. Basically: Why do we need the smart grid now? This should be demonstrated and measured in terms of economics, lifestyle enhancements, energy conservation, and renewable energy integration. The value proposition of the Internet — making the world smaller and connecting every one of us — was more clear to consumers, and the Internet was its own chief marketing officer in its early days.

9 Responses to “A Dozen Things the Smart Grid Can Learn from the Internet”

  1. I’m dumber now for reading this. The title should have read “A dozen things the smart grid can learn from the Internet that the Internet learned from the electrical and telecom industries”. Or maybe this guy should have done some research beyond Google.

    1: Right — the Internet never has scalability (i.e. bandwidth) problems or localized outages. Never had them either, right?

    2: The Internet isn’t OSI model. Incidentally, the OSI model came out of the ITU — a telecom standards body. Yep, old, crusty telecom. The TCP/IP stack kindof maps to OSI, but the OSI is an unimplemented ideal. And if you look into the history of TCP and the Internet, you’ll find it wasn’t always one big happy California love-fest of TCP!

    3: Security was NOT built into the Internet at ANY layer. That came later, in pieces, and patches, and it’s still not there for a good part of it.

    4: “Build web pages?” is your answer for management tools? Really? A better analogy might have been SNMP, if you had to stick to the Internet metaphor.

    5: Please tell me, what are these “open API’s” of the Internet? Again, claim without example. Still, if the author meant perhaps the transport protocols and such defined by the RFC’s, the best ones were written by one or two people who were solving a problem — not by a committee defining a spec for a solution to a problem that doesn’t yet exist in the real world.

    6: This reminds of the Slashdot success formula meme: “1. buy a computer. 2. learn to program. 3. ????? 4. PROFIT!!!!” So, in other words, to be successful, it has to build something successful. Genius.

    7: …. to do what? Seriously, give some example of why I should care how the power I use is routed to my house.

    8: Ahh, task forces. What would we do without ’em? Isn’t this a bit the same as #2 and #5?

    9: Uh, wasn’t this #3?

    10: Same as #1. But, please give an example of an Internet disaster where the same equipment was providing telephony service.

    11: #8 again?

    12: Uh, no. Pornography drove the popularity of the Internet.

  2. Regarding bullet #5 “Open APIs”…
    An “API” already exists for integrating Transmission and Distribution systems and components, including their supporting software. It’s called the Common Information Model, or CIM, for Electric Utilities. CIM is an enterprise model that is vendor neutral and allows disparate systems and components to communicate via an industry agreed upon and built UML model. The infrastructure items mentioned in bullet #5 are easily covered by CIM. CIM has been developed by utilities, for utilities, and is suppoorted by leading electric utility vendors. CIM also works with Home Area Networking models such as Zigbee. Check it out at the URL provided above.

  3. The current state of the grid is so out of date and has not recieved an overhaul in almost 50 years. It’s old technology! In regards to point 4 and 5 I like what you said. It reminds me of a phrase “You can’t manage what you can’t measure” Access to immediate data on your consumption will greatly impact Demand Response, and Conservation Management. Imagine if we open up some of that data to the cloud and what kind of interesting scenarios we could have? Perhaps you’ll be able to log into your twitter feed and see your hour by hour consumption on your mobile phone?

  4. False analogy.

    If anything the smart grid is more like the NYSE, not the internet. It is something that (in theory) can provide greater efficiency through improved liquidity of value of the commodity that is electricity.

    But beware. Here be dragons. Electricity is poorly understood as a commodity that is not regulated.

    Oops. I forgot. Most of the readers of this blog are Californians. You like to de-regulate things like electricity, then run around like headless chickens when everything breaks.