Soon after the New Year, I noted the convergence of protocols, alliances and methods for the Internet of Things was underway.
I know, stop laughing. Industry watchers have predicted this convergence for years, and yet the proliferation of standards and proprietary solutions proceeded each year.
However, in 2016 we saw HomePlug declare victory and transit remaining work to Wi-Sun. Similarly, the Connected Lighting Alliance declared it has accomplished its mission and closed down. The Open Interconnect Consortium and AllSeen Alliance merged to create the Open Connectivity Foundation (OCF).
Companies and alliances will continue to work on many different standards and protocols, but the market is harsh and will sort out winners and losers. Those alliances that do not get sufficient traction can drag on for a long time but become irrelevant to the larger market growth.
Beyond the convergence of some alliances, I see a convergence in the wireless protocols used for low-power devices. The market has been converging on IP-based solutions around UDP, CoAP and JSON objects with CBOR compression. This set of protocols is used in Lightweight M2M, IPSO Smart Objects, OCF as well as the ZigBee Alliances’ new dotdot language for IoT. Alliances like Fairhair are working to provide commonality in the network infrastructure and device bootstrapping for commercial buildings.
Convergence of these protocols is the first step for the market. It’s recognized that for lower power networks and battery-operated devices, existing protocols such as TCP, HTTP and XML needed some optimizations, and the Internet Engineering Task Force has been working for a number of years on these alternatives.
Now many of these groups are developing their own application data models on how, for example, a connected light bulb offers services and interacts with a light switch. In a converged world, there must be agreement on these object data models so a light from one manufacturer running on one technology can be controlled by a switch from another manufacturer on the same or even a different technology.
It helps to agree on what protocols and object models have some actual mindshare in the connected home market. Today shipments in this market are using the following technologies:
- ZigBee has been shipping since 2006, and volumes really grew in 2008-2009 and onward.
- ZWave is used in many home automation systems. ZWave announced in 2016 they had shipped 50 million devices, which is a respectable number.
- Apple HomeKit provides connectivity through iOS but is not widely deployed in the home today. Shipment volumes are hard to determine.
- OCF has been announced, and an open-source implementation and downloadable specification for their home automation solution are available. There is no real use of this protocol in products yet.
- IPSO Smart Objects and Lightweight M2M have been around for several years, but we don’t see large use of the protocols in products yet.
- Thread has been announced, but there are no Thread-enabled products in the market. Thread is a networking stack, and the industry has not agreed upon an application layer to run on it.
- Wi-Fi is used in many products but generally not with a standard application layer that can be used between devices.
- Bluetooth is used in many products but generally not with a standard application layer that can be used between devices.
This list makes it pretty clear the industry has not agreed upon a single application layer. However, the announcement of dotdot helps to clarify this situation.
Dotdot is an IP translation of existing ZigBee device objects. It can run over any IP stack so it can unify existing ZigBee devices with those running over Thread, Wi-Fi or Bluetooth.
While I am hopeful to see the beginning of consolidation of some alliances in 2016, we cannot expect the high-level mergers of groups to align leading wireless protocols. The reality is manufacturers are making choices every day on what technologies and protocols are going to be used in their next-generation of products.
They do not like changing technologies or protocols, and they do not like leaving stranded devices. That means there is a built-in advantage for widely shipping technologies like ZigBee and its migration to dotdot.
If this can be done quickly and deployed to market, it can help to consolidate not only the underlying protocols in use, but also the object data models within the devices. Then we can have simplified interoperability and connectivity to our cloud and mobile devices.
--Skip Ashton is vice president of software at Silicon Labs. He is a member of the board of directors of the ZigBee Alliance, the Thread Group, the Connected Lighting Alliance and the Fairhair Alliance.