Back in June, the Thread Group unveiled version 1.2 of its namesake protocol, marking a shift of sorts to focus on commercial buildings as well as smart homes. A few months later, the Google Nest Hub Max smart display became the first publicly available ‘Built on Thread’ device, which was definitely a bit of a late arrival.
We had a bit of a rant back in August, when Zigbee announced that it had now certified 3,000 devices, and were a bit despondent about the state of this low-power personal-area-network (PAN) market, where we did argue that Thread should just hurry up and merge with Zigbee properly. But touching base with Thread Group president Grant Erickson, it seems clear that, if anything, Thread wants to integrate with as many application layers as possible – in its new drive to tackle the commercial smart building sector.
Erickson said that the reception of v1.2 was within expectations, and noted that companies can start with v1.1 and upgrade fairly easily, to enable the new features. The three major additions consist of upgrades to the low-power enhancements, which Erickson says that in the Nest Detect Window and Door sensor can add 50% to the battery life.
The second major pillar are the new Bluetooth Extensions, which Erickson says allows you to standardize the commissioning of Thread devices using Bluetooth – such as when adding a new smart home device to a home via a smartphone, as the out-of-band communication. You can now also virtualize a Bluetooth-only device as a Thread device, using the extension, which is a way to bring IP capabilities to such devices and treat them as part of the same single network.
The final addition are the Commercial Extensions, which Erickson said are focused on catering for the differences between how you would deploy and manage a home network versus one in a commercial environment. The latter requires multi-user administration and permissions, as well as a way to treat multiple Thread networks as a single network. Erickson did note that there is a retail deployment in France that has 1,800 nodes within a single network, which is rather a lot more than the conventional recommend practical limit.
Of course, we put a few difficult questions to Erickson, chiefly that progress in Thread adoption has been slow and that Google’s Nest Hub Max being the first ‘Built on Thread’ certified device was emblematic of the politics at play. Erickson countered that Apple’s addition to the board of the Thread Group should be noted, given Apple’s historic avoidance of standards organizations until it has decided on a concrete product strategy.
Similar things can be said of Lutron, another recent addition, and Erickson said that Lutron had carried out a bakeoff between the different wireless options over the past two to three years before deciding on its horse in this race. With Apple, one would expect Thread to start appearing in the HomeKit ecosystem soon, but that platform is a long way off the pace compared to Google and Amazon.
As for timelines, Erickson said there would be announcements in the coming months, and that 20 to 40 devices are likely to be certified in the next year. As for the group’s internal activity, it is split equally in its efforts on the smart home and the commercial and enterprise spaces, with the two sectors having a lot of overlap – in IT/OT convergence and the move to end-to-end IP addressability.
On the matter of Thread frequently using the Zigbee application layer, Erickson said that the Thread Group now has demonstrations of a single Thread network running using 5 simultaneous application layers, including WiFi, Google’s Weave, Zigbee’s dotdot, KNX-IoT, and BACnet. OCF’s IoTivity Lite is also supported here.
Sure, having a plethora of application layers to choose from means that there’s a lot of flexibility in the schema, but our view has always been that it is another potential roadblock that a developer will run into. Having a single application layer to flash to all Thread-enabled devices would simplify development, but it is Thread’s view that the flexibility is needed.
We asked Erickson if the Thread Group was worried about Bluetooth Mesh, another protocol that has taken a long time to reach the marketplace. Erickson quipped that despite the $35mn marketing budget, there wasn’t much to worry about here, and that the general consensus is that by the time Bluetooth sorts itself out, there are going to be new versions of its low-power mesh rivals. He added that the need for an IP gateway for Bluetooth is a burden that Thread does not have to bear.
As for his thoughts on the marketplace dynamics, Erickson says that there’s still a lot of frustration in the consumer space, but that the arrival of the digital assistants in the Google and Amazon ecosystems have helped solve the problem of device integrations.
Thread is, of course Erickson notes, thrilled that Google is an ardent supporter, and when we asked about thoughts on Amazon, Erickson said that while Amazon is quite entrenched in the Zigbee ecosystem, and Ring is firmly in the Z-Wave camp, the two ecosystems do have a shared base in IEEE 802.15.4 radios, and that there is nothing precluding Amazon from pushing software upgrades to add Thread support down the line. Erickson added that the two camps realize that they are fighting over a small slice of a total market, and that everyone is going to have to make some changes, but stresses that a rising tide floats all boats.
We closed by asking for Erickson’s thoughts on the smart home market as a whole, where he noted that the sector had grown much more slowly than most in the industry would have hoped. However, he is still excited, and recalls the modem wars in the 1980s that many parallels can be draw to. It took the emergence of the World Wide Web as the killer application to bring an end to that Wild West period, and in the smart home, there is going to have to be a similar transition.
We asked if this meant that a single protocol had to emerge as the victor, but Erickson thinks that end-to-end IP addressability is the key to proper convergence. There might be some consolidation among the application layers, but as long as you can send IP-based messages, you can afford to have multiple link layers in a deployment. Erickson said that it was clear that the smart home market has many multi-billion dollar segments, but that the challenge is moving up from the plumbing to the porcelain.