The Battle for Mesh. Bluetooth vs Thread?

It’s a New Year, which means it’s time for the annual week of madness in Las Vegas which is the Consumer Electronics Show.  For four days, the electronics industry comes together to tell consumers what they ought to be buying, whilst analysts and the media try to predict what will really be the hot product sector for the coming year.

Over the last few years, as PCs, tablets and phones have lost their wow factor, that’s proven to be a little more difficult than it used to be.  In 2014, the consensus was that wearables would be the next big thing.  They have definitely made strides beyond basic step counting, but are still smouldering rather than setting the world on fire.  Instead, the innovation which caught the public imagination at CES in 2014 was the selfie stick.

In 2015, the smart money was on smart homes.  But with a few exceptions, consumers felt the smart thing to do with their money was to buy more selfie sticks.  This year, the pundits will probably predict that 2016 will be the year of the drone.  My guess is that most consumers will still prefer to buy selfie sticks.  Unless someone comes up with cheap drones that take selfies.

Of course, like all good works of fiction, the CES show contains a number of interesting subplots,  one of which will be the battle for mesh.

Mesh is an interesting topic.  It’s not new – we’ve had ZigBee and Z-Wave for over a decade.  But despite the amount of money that companies have spent developing the different standards and launching products, they’ve still not made much of an impact in the consumer market.

Last year, two heavyweights announced new mesh technologies which they both claim could change that.  Shortly after CES, Google and Nest announced their own mesh network, called Thread.  It was a “clean slate” development, supporting an IPv6 mesh, developed by a small group of companies which had extensive experience with mesh in the ZigBee world.  Then, in November, the Bluetooth SIG took an unusual step in revealing the contents of a future specification release, announcing the imminent launch of Bluetooth Smart Mesh.  It’s not the first Bluetooth mesh – CSR, prior to their acquisition by Qualcomm, had introduced their own proprietary mesh protocol at CES in 2013.  However, the SIG announcement indicated that the group was actively developing a low power mesh standard which would be interoperable.  Neither Thread nor Bluetooth’s mesh are compatible with ZigBee or Z-Wave products, so they both need to convince manufacturers to renounce their current choice of mesh standard and join their respective camps.

The inevitable reaction to both announcements was a spate of speculation about which would win the battle for mesh.  That speculation is likely to ramp back up, as both camps will have companies lining up to announce their respective support at CES, setting battle lines in the “my lightbulb’s smarter than your lightbulb” war.

In fact, the battle is elsewhere.  Neither the Thread nor Bluetooth mesh standards are likely to be ready for mass consumer deployment in the next few years.  What they will both be aiming to do at CES this year is to pull the rug from under the competing mesh standards of ZigBee and Z-Wave.  Smart Home has not yet become a commercial success for a number of different reasons, including cost, compatibility and an inability to persuade most users that it’s an improvement on their current non-smart home.  That situation is unlikely to change much in the next few years.  However, despite that lack of consumer demand, both Thread and Bluetooth are perceived by analysts to have come into the market late, which means that in the short term the most important battle for both of them is to sow confusion to stop any chance of ZigBee or Z-Wave succeeding.  It doesn’t really matter to either of them if the smart home market is delayed for a few years, so long as nothing else establishes a dominant position.  So the name of the game is “trash the incumbents”.  I suspect they may do that rather well.

Mesh is frequently misunderstood, or given more credit than it’s due.  By itself, it is just an enabler.  Its primary purpose is to extend the range of wireless devices without the need for them to increase their transmit power.  As power consumption rises steeply with output power, that’s very important if there are going to be large numbers of sensors around the home.  The presence of lots of mesh devices provides the wireless “fog” infrastructure that can connect everything in the home, ensuring that you can make connections through walls and between floors.  That’s it.  Mesh doesn’t provide the language for devices to converse, or define what they say to each other.  Smart home proponents are often a little foggy about this concept, missing the point that these interactions need to be defined and developed.  You can’t just assume that devices will magically find ways to have meaningful conversations with each other, making themselves more efficient and thereby adding to the overall sum of human happiness.  Many smart home advocates ignore this fact and sound as if they’re the technological equivalent of Doctor Doolittle:

“If we could talk to appliances, man to microwave,

Maybe take a dishwasher degree.

I could banter with my beater, hobnob with my heater,

Think how smart my little home could be.”

 Beyond some smart lighting and HVAC control, most of what the industry is trying to sell consumers still feels very speculative and fails to convince people that it’s an advantage over their current dumb home.  Most scenarios involving my door lock talking to my lightbulbs, washing machine, toaster or dishwasher just seem contrived; more likely to complicate my life rather than making it easier.

There was a very good article recently in EE Times, where Skip Ashton was interviewed.  Skip is the VP of Technology for the Thread Group and one of the gurus of mesh who probably has more practical experience of it than anyone else.  He pointed out the real problem regarding the language of appliances.  Mesh may give them the capability to talk to each other, but that’s all.  It’s like being able to phone a number in China.  It’s easy to dial the number, hear it ringing and get an answer.  But if you don’t speak Chinese that’s not necessarily a lot of use.   And even if you do, it’s still pretty pointless unless you’ve got something to say to the person at the other end.

Whilst there are only have a handful of mesh standards, there’s an order of magnitude more organisations and individual companies trying to develop application level languages which sit above the mesh protocol and let smart home devices talk to each other.  Most still don’t appear to have much idea of the resulting use cases – it’s as if they’re writing a pocket dictionary in the hope that once it’s published a Shakespeare or Dickens will inevitably emerge.  History suggests it needs to be the other way around.

Industrial mesh applications are doing a lot better, but that’s because they don’t have the same need for interoperability.  They’re driven by specific requirements in a particular sector, which can be shown to give a return on investment.  In most cases they’re provided by a single supplier who has a deep application knowledge for that sector, whether it’s agriculture, warehousing, retail, etc. so there’s no need for interoperability as each is effectively a proprietary solution.  Everyone knows what the devices need to do and what they need to say to each other. 

In the consumer world it’s a different game, where manufacturers are largely selling products to achieve critical mass, in the hope that at some point the compelling applications will emerge.  That does require interoperability, hence the battle to be “the” smart home application standard. That’s where the real battle is, not at the mesh layer.  It’s only just starting and won’t be won for many years.  In the meantime, smart home devices will probably continue to be largely proprietary, so you’ll be taking a risk if you mix lightbulbs or thermostats from multiple vendors.  But that may not matter, as no-one’s really worked out why many of these devices would ever want to be connected.  By the time anyone’s sorted that out I expect we’ll see Thread and Bluetooth coexisting nicely, or even coming together in a future marriage.  However, right now, they have a common, vested interest in spiking any other mesh technology. 

So over the coming months, expect to see their supporters generating the maximum number of column inches to imply there’s a real battle between them, rather than just a phoney war.  Although that’s usually what CES is all about anyway.