lousy or changing RF links or by having too many good links and therefore too much chatter on the network. We think about these problems much more holistically than guys who started out as chip companies.
The smart-energy guys also wanted high security with public keys, since there’s real money [such as utility bill rebates] riding on how these networks work. So we put in security and a bunch of other things to facilitate messaging and now the network responds.
We have put in certain elements of our IP that we thought were necessary to make this world-class. But it’s not like we’re getting royalties—we have donated to the cause of having an open standard. With Bob Metcalfe as our chairman and first investor, he has lived the fact that open standards win in the long term. We are not a ZigBee company, that is not all that we can do, but that is what we have standardized on.
X: Stepping back a bit—there’s an obvious need for more networking between devices in homes and other facilities, but why does this networking need to be wireless?
RL: For a green-field building, depending on how you do your installation, you can make an argument that installing wireless systems is simpler and therefore cheaper. And if you’re coming into an existing facility, then there’s no question that wireless is preferred to cutting through the walls and running wires. The other thing is that in most facilities and homes, people tend not to monitor the things that you can’t run a wire to. But now, putting an extra thermostat in the baby’s room becomes that much simpler.
The only issue is that wires are still the gold standard. You know that if something has a wired connection, the wire is not going to be your point of failure. So you need to make sure [a wireless system] is robust.
X: And if you’re going with wireless, why should you go with ZigBee? And why a mesh networking approach, rather than some other protocol?
RL: If you’re doing a smart-energy application, you don’t want to be wasting energy; you don’t want to be using half a watt to control a 20-kilowatt device. If you think of a graph with power requirements on the Y axis and speed of communication on the X axis, Wi-Fi is in the upper right at 50 megabits per second and high power consumption, Bluetooth is right in the middle at 1 megabit per second and medium power consumption, and we and ZigBee are in the lower left at 250 kilobits per second, with a typical ZigBee device running for four or five years on two AA batteries. We’re always driving down the power requirements.
The reason for mesh networking is that a Bluetooth network can only handle about seven devices. For point-to-point communication, low-power Bluetooth might be a good solution. But if you want something to live inside a broader network with tens or hundreds of devices, then you need to have a mesh, not just for reliability, but for the ability to communicate at all. Message arbitration and management is an inherent part of the ZigBee standard.
X: How will these wireless smart-energy devices take hold first? What will they do?
RL: In the U.S. we’ll see smart meters first—that already started in 2008, with California and Texas taking the lead. You will see some large-scale pilots in homes this year at relatively low volume, and ramping up in 2010. Your typical electric meter will now be a solid-state device, a mini-PC rather than an electromechanical device. That meter will have two radios in it—one for two-way communications between the meter and the utility, and one going from the meter into the home, which will be the Ember-ZigBee one. These will be replacing the electric meters in the majority of residences in California and Texas, but that will be a 3- to 5-year effort.
Inside the home, you’ll have displays showing electricity usage and pricing information. They may have a “red,” “green,” or “yellow” signal depending on whether energy is cheap or expensive. If you are part of a utility program, they will have the ability to increase or decrease your thermostat remotely. It may also show you the energy being consumed by specific appliances [information collected by Ember-equipped “smart plugs” around the house]. The utilities are looking at aggregating this information so you can see your carbon footprint compared to your neighbors’; they’re getting very creative about what you will be able to do. But the biggest thing is to able to see the full usage for the month, so you can go in and see how much energy you’ve used, and how you’re doing on your bill.
X: I guess the theory is that if you give energy consumers all this information, they’ll automatically make better decisions about how they use energy?
RL: Today, buying energy is like going to a supermarket where there isn’t a price on anything, you have no idea how much you’re buying, and a month later you get a bill saying, ‘This is how much it cost you and this is how many calories were in your basket.’ That is what we have been tolerating in the energy market. So yes, one scenario says that all you need to do is give people information and they’ll do the right thing. But we can’t plan new power plants on the hope that
