Why We're Ditching ZigBee - But Keeping The Good Bits

Posted : 21-11-2012

It's been a difficult decision, but we've decided to stop further development of our ZigBee compatible products. That said, we're still planning to do some pretty interesting things with 802.15.4 based wireless mesh networking.

At Zynaptic Limited we've been working with ZigBee technology for over 6 years, and recently we have been carrying out significant development work on our own ZigBee based products. However, having evaluated the costs and benefits of joining the ZigBee Alliance we have come to the conclusion that it is no longer in our best interests to sign up as members.

The ZigBee Alliance : What Went Wrong?

It is now over five years since we were privileged to attend the Grenoble ZigFest (under the auspices of AlertMe). At the time, the ZigBee Pro specification was nearly complete and the networking stack vendors who attended were addressing the final interoperability issues. It was obvious that a significant amount of effort had gone into solving the difficult technical problems associated with low power mesh networking, and there was a general air of confidence that the new standard represented a solid base on which to develop a range of new and innovative low power wireless products.

If we had been making our decision then, we would have had no hesitation in joining the ZigBee Alliance. Unfortunately, the nature of the organisation has radically changed in the intervening years - and we no longer feel that it is representative of our ambitions as a business. For the backstory behind some of the changes we would recommend reading this article by Bob Gohn at Pike Research. The bottom line is that our perception of the ZigBee Alliance is now that it is less of a useful standards body and more of a political football to be kicked around between whichever companies are prepared to pay the $50,000 annual entry fee.

What To Keep And What To Throw Away

We know that we are not the only developers who have become disillusioned with the direction that the ZigBee Alliance has taken, and there are already a number of independent efforts to create alternative technical solutions. However, our feeling is that discarding all the good work carried out in developing the original ZigBee Pro platform is hugely wasteful - particularly given the quality and maturity of the various networking stack implementations that are already available. The question that must be asked is whether it is possible to take advantage of this legacy platform and to continue developing it independently from the ZigBee Alliance.

In our opinion, the main failing of the ZigBee Alliance over the past few years has been its inability to formulate a single coherent approach to application design. Instead there are a rapidly increasing number of vertical application frameworks - each of which is championed by a different faction within the ZigBee Alliance. These application frameworks tend to be mutually incompatible, but with significant functional overlap whenever two different factions disagree on the best way to solve a particular problem. A prime example would be the duplication in functionality between the existing Home Automation application profile and the new Light Link application profile.

We have now decided to carry out further development of our products using an independently developed and openly documented application profile running on top of a well defined subset of the Silicon Labs EmberZNet Pro networking stack. Observant readers will recognize that the EmberZNet Pro stack is actually a fully qualified implementation of the ZigBee Pro specification. However, since we will not be using it in a manner approved by the ZigBee Alliance our products will no longer be classed as ZigBee devices. Instead we will refer to our device networks as using the 'Buzz' protocol - since this is what bees should be doing when they 'just work'.

Getting With The Buzz

One principle we want to adhere to when developing our products is that every design decision should be driven by what is best for the end-user. This is particularly important when it comes to ensuring that the end-user has complete control over their own data and a free choice when it comes to selecting any online 'Internet of Things' services they want to use. This means that our initial focus is on trying to get a beta version of the technology into the hands of any developers who share these aims. If you fall into that category, please keep an eye on our Twitter feed (@zynaptic) as we will be providing a lot more technical information in the near future.