The Smart Metering and Smart Homes exhibition in Amsterdam is Europe’s largest show for this industry, so it’s a good reflection of where things are at. Given the amount of noise that’s been generated around ZigBee and its Smart Energy Profile, I’d expected to see most of the other wireless contenders to be absent or skulking in their cages. However, someone walking around without those preconceptions might have gone home with a rather different view of the state of play.
There’s no doubt that ZigBee is well placed in current smart meter deployments. Although there are quite a limited number of real ZigBee deployments in Europe, the UK has more or less committed to SEP 1.2 for its foundation phase of national deployment and most meter and IHD suppliers were showing ZigBee products, albeit with not very many sporting a ZigBee certified logo.
Despite that, a significant number of suppliers were also highlighting support for the new Wireless M-Bus standard, which has slithered down the spectrum to its new resting point of 169 MHz. Wireless M-Bus has always had a popular following within Germany, with an implementation based on a radio running at 868 MHz. The shift to the lower frequency acknowledges one of the enduring complaints which the 868MHz camp has levelled at 2.4GHz solutions, which is their potentially limited range.
Whilst 2.4GHz is a frequency that’s fine for most houses, it faces challenges with blocks of flats. Up until now, the 868 MHz triumvirate of Wireless M-Bus, Z-Wave and wireless KNX had always given the impression that they could achieve adequate range at 868 MHz. This break in the 868 MHz ranks does not augur well for a reasoned debate, but just increases the in-fighting and paranoia about whether any radio standard works or is ready for deployment. That’s not what Smart Metering needs.
Lower frequency proponents have a history of playing mischief on the 2.4GHz standards, citing their better range. Whilst that’s technically true, it can be a specious argument. Moving away from 2.4GHz brings other problems, not least of which is the lack of bandwidth in the lower ISM bands. That limits the throughput which can be achieved by operating at a lower frequency. It may not be a major issue for transmitting measurement data, but it means that features like Over the Air upgrades may be challenging.
The PR around these lower frequencies solutions generally focuses on the range of the radio, but that’s only part of what you need for a smart meter or smart home deployment. The bigger development task is the protocol stack which brings security, power management, topology and commissioning tools to the solution. A lot of the lower frequency standards jostling for position have relatively simple stacks. In one sense, that’s good, as there’s nothing wrong with simplicity. But for an application like smart metering, where security and integrity of data are paramount, these simplistic approaches look rather limiting. The industry needs to understand that the time for talking and pontification is over. We’re only really going to discover what works by committing to deployments, rather than pandering to academics and consultants who want to be paid for telling us there’s something better around the corner.
Wireless M-Bus was not the only contender. The main Smart Metering show had outgrown its space and spawned a daughter exhibition and conference targeting the Smart Home. On the exhibition floor for Smart Homes, there seemed to more floor space taken up by wireless standards that there was devoted to products. Z-Wave were proudly demonstrating the range of products that they’ve enabled, as was EnOcean. Wireless DECT, a recent latecomer to the table were busy promoting themselves, as were wirelessKNX – another favourite in Germany. In comparison, the ZigBee booth looked rather small and apologetic.
Around the show floor, there was the initial evidence that ZigBee may have let in a Trojan Horse by embracing IP in their Smart Energy Profile 2.0. I’ve written about the incompatibility of IP with low power radio before and my personal view is still that it’s a really bad idea. But the IP lobby has gained momentum, especially since the decision to allow it to run on different radios. That was evident in Amsterdam. SEP 2.0 is still a long way off, but I came across four separate stands with demonstrations of an early implementation. What was interesting is that only one of these four ran on a ZigBee radio – the demo being given by Ember. The other three were all using Wi-Fi. And as far as I could tell, these were all using stacks from different vendors, running on different host microprocessors. So a lot of effort is being put into SEP 2.0 by people who have no interest in ZigBee. It’s quite clear that the Wi-Fi community see this as their chance to wrest smart metering and home automation away from ZigBee.
For those with a long memory, ZigBee isn’t the first wireless standard to get mugged by the IP fundamentalists. Back in 2002, when Microsoft was starting to court Bluetooth, it led a similar crusade to incorporate IP support into the Bluetooth standard. At conference after conference they would present the world according to IPv6, to the extent that it became a favourite game for delegates to lay bets on how many seconds it would be before the Microsoft speaker uttered those four syllables. The IP lobby got sufficient traction for many man years of development work to be undertaken, some might say wasted, leading to the PAN profile, the BNEP Network Encapsulation Protocol and an Automation profile proposal using BNEP to support ANSI 1451.5. All of which have subsequently been ignored by product and application designers. The only significant think to come out of this IP invasion was that its major proponent became the Chairman of the Bluetooth SIG. So maybe there’s a lesson for Bob Heile (Chair of the ZigBee Alliance) to keep a close look out for Geeks bearing IP gifts.
To some degree, it doesn’t matter too much which standard is chosen for any national or regional smart metering deployment. Utilities have always tended to purchase proprietary versions of meters and back ends – a fact that has led to the observation that US utilities may be wasting over $2 billion is their procurement policies. So if the UK goes ZigBee 1.2, Germany goes wireless M-Bus, France goes Powerline and Massachusetts goes Wi-Fi SEP 2.0, that will work for utilities. What it does not do is open up a market for connected home automation devices. The point here is that whereas utilities determine exactly what they put in your home, when it comes to appliances, it’s the consumer that chooses. Which makes it is really important for the industry to coalesce on one single radio standard. And I emphasise radio, as it’s the physical layer that is important here. Wireless standards are complex, which means they are both difficult and expensive to implement. They’re also difficult for consumers to commission. Most appliance manufacturers are not familiar with them and don’t want to have to support multiple options for different markets.
Today you can buy some top of the range domestic appliances with Wi-Fi, Bluetooth or Z-Wave, but they’re expensive prosumer products for the geeks. A lot of the speakers in the smart home conference were promoting a vision where you can lounge in front of your TV and use your smartphone or tablet to turn off the annoying bleep on your dishwasher. (In case you hadn’t guessed, the utilities are starting to recruit marketing managers from the telecoms industry.) Whether or not you believe in that vision, it’s only going to start to become a reality when the industry decides on a wireless standard that spans phones, TVs and dishwashers. There is a school of thought which believes that we’ll buy gateways that will translate between all of the different wireless protocols. I don’t buy that one. I’ve heard it before, as a solution to bridge between different combinations of Bluetooth, Wi-Fi, ZigBee, DECT, Insteon, X-10 and IrDA. These gateways exist, but they’re expensive and complex to use. I see nothing that will change that.
What the industry needs is a single standard to coalesce around. That will drive volume, which will reduce price, which will increase its penetration from high end to standard appliances. There’s lots of individual alliances and demonstrations going on, but they’re still at the stage of teenage groping, not meaningful, long-term relationships. It doesn’t help that Government sponsored bodies, such as NIST and the EU are trying to impose their own agendas, like prudish parents who are attempting to enforce an arranged marriage.
Today the assumption is that home automation will follow the lead of smart meter deployments, as these will be the primary source of consumption data. However, that may not be a valid assumption. The utility industry has its own, peculiarly slow deployment timescales and a risk aversion that means it will probably fight to retain control of data. Each utility or national Government may well mandate their own choice of wireless standard. But by the time it’s deployed, the consumer industry may have gone down a different route. Nest Labs has just launched its Wi-Fi thermostat. It’s not the first, but the company’s founders have learnt lessons from their time at Apple about how to make a product desirable. It’s certainly going to raise the barrier for home HVAC products.
But back to Amsterdam. One of the most bizarre aspects was the closing session of the Smart Homes conference, which saw a panel session that crammed the stage with representatives of most of the competing standards. The ZigBee Alliance was there, along with HomePlug Power Alliance, the KNX Association, Ultra Low Energy DECT, the OSGI Alliance, EnOcean, the Home Gateway Initiative and Z-Wave. All was sweetness and light as they agreed that there was room for all of them to work together. But you got the distinct feeling that you were looking at animals at the Zoo who were waiting for night to fall, each in the knowledge that their keepers had forgotten to lock the cages. And each of them knew that as soon as the lights went out, one or more of them would be on the menu.