Embedded Computing Design, Brandon Lewis, January 13, 2017 – I’m sure you’ve all seen the following comic before and chuckled when thinking about the proliferation of Internet of Things (IoT) standards.
Until recently, it seemed like every month or two a new “IoT standards organization” was popping up with the promise of application-layer interoperability for connected devices. These bodies often flew under the flag of “open source” to varying degrees, and were almost certainly always championed by industry giants that waived an invisible hand over the development and direction of the “standard” while the other members half-heartedly participated or joined simply to take a flier on what could potentially be the leading implementation of choice. As a result, an ironic climate of competition developed between the various standards bodies that were said to be working towards unity and compatibility, with groups such as the Open Interconnect Consortium (OIC) with IoTivity, AllSeen Alliance with AllJoyn, Thread Group, IPSO Alliance’s Application Framework, Open Mobile Alliance’s Lightweight M2M (LWM2M) protocol, and a host of others all looking to gain traction, and thus, developer mindshare. It seemed, at a point, there might even have been more standards bodies than actual standards.
However, developments over the last six months have started to reverse that fragmentation, as AllSeen is now all gone after merging with the OIC to form the Open Connectivity Foundation (OCF), and others are beginning to fade into irrelevance. Furthermore, complementary organizations are now collaborating on software layers that will enable devices compatible with their respective technologies to interoperate, with one such example being the dotdot language announced by the ZigBee Alliance at CES 2017, and supported by the Thread Group.
dotdot: Wireless standards unite
In essence, the dotdot language can be seen as an extension of the ZigBee Cluster Library (ZCL) specification used to issue commands across ZigBee 3.0’s interoperable application layer. In addition to the application layer remaining slim, however, dotdot has also been made compatible with other networks to allow for communication across disparate transports, with Thread being the first non-ZigBee qualified network topology for dotdot certification thanks to a liaison agreement between the two organizations. Therefore, ZigBee-based devices will be able to communicate with IP-based devices residing on a Thread mesh network, with other transports scheduled for certification over the coming year (Wi-Fi, Bluetooth, other?).
“[ZigBee Alliance] members wanted to use the ZigBee application layer [for dotdot] because it was universal across manufacturers, as well as ZigBee’s already established ecosystem,” said Neal Kondel, one of the leaders of the dotdot initiative at the ZigBee Alliance.
The implications of dotdot are significant for both developers and end users, as the former will be able to potentially maximize code reuse rather than optimizing their designs for different networking technologies, while consumers will benefit from not having to choose between connected devices that will result in vendor lock-in. As Kondel mentioned, because the ZigBee ecosystem is already well established and the dotdot language is already compatible with existing ZigBee devices, the rollout of dotdot-based products should be accelerated.
The goal is universal, as dotdot intends to be “a generic solution that becomes a conduit for all other networks,” Kondel added.
In Thread news, the organization is currently in the process of certifying several vendor implementations for compatibility with the Thread protocol, including ARM, Silicon Labs, and NXP. Grant Erickson, President of Thread Group expects these products to achieve rapid market adoption in the later part of this year and into next, with future activities planned around market segments such as the enterprise and commercial sectors.
This growth, combined with the dotdot language, should prove to be a step in the right direction for IoT interoperability.