ZigBee finally arriving?

electron

Administrator
Staff member
There has been a lot of news about ZigBee lately, so I figured I would post some of the more interesting articles I am finding on the internet.

Dev kit:
MOUNTAIN VIEW, Calif. & SURREY, England--(BUSINESS WIRE)--Sept. 18, 2006--Integration Associates:

-- Based on IAI's CompXs Technology, Integration Associates (IAI) Today Announced Their Pre-Certified USB Dongle Developers Kit Designed to Shorten the Time-to-Market for Designers Building 802.15.4 / ZigBee Products

Integration Associates, manufacturer of low cost, high performance analog solutions for wireless and wireline data communications, and of the OEM 802.15.4/ZigBee USB Dongle, has today further expanded its EZLink(tm) wireless product line to include a USB Dongle Developers Kit for 802.15.4/ZigBee product development. The EZLink product line is aimed at OEMs looking to reduce R&D cost, reduce time-to-market, and minimize development risks.
Source

PC World got a briefing on Zigbee home automation technology two or three years ago, and it sounded very promising. But then it fell off the face of the earth--we've never seen any Zigbee products. That's about to change; the organization behind the technology says we'll see "tens" of Zigbee-certified consumer products by the end of the year.
Source: PCworld
 
Sometime over the next day or so I'll post some of my notes from CEDIA. Well except for the ones that might get me sued...

I will say this about ZigBee - Not much has really changed on the product front. Control4 is still one of the few companies actually shipping anything in volume that looks like ZigBee.

Vantage Controls was showing their "ZigBee" lighting products and I couldn't get anyone to speak in detail about them. I did find out that the sales model is business to business which equates to low volume in my book. Plus Vantage was just sold to Legrand which throws some uncertianty into the future of the product line.

The Eaton Home Heartbeat is truly cool and also looks real. Shipping through the builder channel now and is supposed to be sold through CompUSA and HomeDepot.com Decemberish. Sensor only though with limited control capability except for the actuator to turn off your water in resonse to a leak detection. I would buy this product for the leak detector/water shutoff feature alone. If you haven't had a chance to see it the industrial design is some of the best anywhere in the industry.

Two comments on the Brent Hodges Quotes in the PC World article:

"Hodges says that the Zigbee specification was ratified in 2004; the reason it has taken so long is that Zigbee is an open standard... with all products made and sold by one vendor: SmartHome"

Brent should have had someone walk over to the Broan-Nutone booth at CEDIA and check out the ventilation switches implemented on INSTEON. Shipping into the builder channel now.

In 2004, right around the time the standard was ratified WTRS was told by the ZigBee alliance that 2005 would be the year of ZigBee and that all of the companies involved in the standard were jockying for position to be the first ones out with products. Then we were told 2006 would be the year. Now we are told that 2007 will be the year. One of these years they may actually get it right.

Maybe Brent, it is unclear:

"Some "Designed for Zigbee" products are already available; they haven't finished the certification process, but they should work with other Zigbee products, and they may be certified later."

As it was explained to us, this certification means that the product uses a processor/radio combination that is capable of running ZigBee and that the network software has been certified for ZigBee. No guarantees about the applications themselves. Basically the applications won't interfere with each other and could be made to interoperate someday in the future.

I ended up talking to the ZA folks at CEDIA much longer than I had planned with the end result that I didn't get to talk to some companies. I didn't come away with enough new information to change my opinion about how soon we will see ZigBee products in any volume in the HA world.

More later.

George West
 

upstatemike

Senior Member
I used to live in Surrey and it never struck me as the sort of high tech environment that could produce something as sophisticated as a "Dongle Developer's Kit".
 

Paul

Active Member
n24402189_30625917_7620.jpg
 

rocco

Active Member
George West said:
As it was explained to us, this certification means that the product uses a processor/radio combination that is capable of running ZigBee and that the network software has been certified for ZigBee. No guarantees about the applications themselves. Basically the applications won't interfere with each other and could be made to interoperate someday in the future.
Thanks for the info, George.

My understanding on certification is different, though I'm not clear on all of the details. Here is what I think I know about the three levels of certification:

Much of the certification that has taken place so far is at a level that is referred to as a "ZigBee Compliant Platform" (ZCP). This is undertaken by the chip-set vendors, like TI, Freescale and Siemens, to insure that the silicon meets the more general layers of the standard. As a product developer, I could avoid certification testing at these lower levels by basing my design on a ZigBee compliant platform.

The middle layers of the protocol, around the network layer, would need to be certified as being "ZigBee Network Capable" (ZNC). This means that it can correctly participate in a ZigBee network, but not necessarily inter-operate with other manufacturers. It is assumed such a device is using a "non-Public Application Profile", most probably because it doesn't fit into one of the "standard" ZigBee public application profiles.

A "ZigBee Alliance Certified Product", which also allows use of the ZigBee logo, is a product that has been certified all the way up to the ZigBee alliance application profile. Application profiles include "Home Control Lighting", "Commercial Building Automation", "HVAC" and "Home Automation". Products certified at this level should inter-operate with devices of other manufacturers.

Whew . . .

Now, even though the ZigBee specification was certified in December of 2004, I have yet to see a ratified application profile (might be just me). I believe this is what is holding everything up. The profiles needed to be written, testing procedures defined to insure inter-operability within those profiles, manufacturers have to design products that they hope will pass those test procedures, and then they have to get into a queue to have the testing/certification done.

Think about the difficulty of writing a profile. It must be written such that every company that tries to implement it interprets it the same, without any leeway that might introduce incompatibilities, whether they are reading the spec in English, Korean or French.
 
rocco said:
Whew . . .

Now, even though the ZigBee specification was certified in December of 2004, I have yet to see a ratified application profile (might be just me). I believe this is what is holding everything up. The profiles needed to be written, testing procedures defined to insure inter-operability within those profiles, manufacturers have to design products that they hope will pass those test procedures, and then they have to get into a queue to have the testing/certification done.

Think about the difficulty of writing a profile. It must be written such that every company that tries to implement it interprets it the same, without any leeway that might introduce incompatibilities, whether they are reading the spec in English, Korean or French.

Hi Rocco,

That is an excellent job of listing all the different ways that a product can be ZigBee certified. Much better than my lame effort. Anyone that's wants to understand this subject can scroll up and read what Rocco wrote.

I was trying to describe the ZigBee Network Capable platform. A company can get this certification today with the option of submitting the application for full ZigBee certification at a later date. Basically a company would transfer over the IP related to the application to the Alliance as part of this certification process. I got the under on that bet.

On my more cynical days, I think this proliferation of certification levels is in response to negative comments about the ZA allowing the definition of ZigBee to fragment in order to claim that any application based on IEEE 802.15.4 was a ZigBee application. There was a lot of noise late 2005/early 2006 about "ZigBee, formerly 802.15.4" from the ZA. A bunch of people pointed out that the ZA was allowing some of its member companies - the ones selling real products - to claim that their products were ZigBee when in actuality the products were running on ZigBee capable hardware, uncertified network stacks and with uncertified application profiles.

AFAIK, the only ratified application profile is a light switch. About two years ago, one of our customers asked what I thought was a very good question: "Since the most common means of communicationg between embedded processors is serial I/O of some kind, what about an SIO profile?" The obstacle to more application profiles is always given as "we need to take the time to get everything just right." I think the main obstacle is IP transfer and sharing.

To pull this back to HA, IMO there is still a considerable ways to go before ZigBee can be considered a viable platform to deploy any control application without also requiring a significant level of post-sales integration and installation support. Sensor applications like the Home HeartBeat? Yes. A product that requires a bounded response like a light switch, and can be sold directly to consumers? Not today.

As much as the ZA likes to bash Zensys/Z-Wave, other than the (Chris is correct should be 232. I keep mixing this up with the capability of one of the remotes) node limit, ZigBee has many of the same issues. When you talk to real people doing real projects using ZigBee you quickly find that the size of the network that can be easily deployed is on the order of tens of wireless nodes. This is a long way from letting you unbolt your ELK or LightJet and replace it with a completly wireless infrastructure.

I read the AMX announcement and can't figure out how they are able to use the ZigBee label on that product without qualifying it as either ZCP or ZNC.

The problem faced by the ZA is that the only identified application that gives them the volumes to make the technology interesting to a TI or the venture community is HA. From a strategy standpoint this means that they need to slow down the adoption of competing technologies (UPB, INSTEON, Z-Wave...) before they reach significant volume and can't be easily displaced. The funny thing is that if the ZA had just got on with business and not spent so much time dinking around with alliance politics we might be seeing more/better HA products using that technology today.

George West
www.wtrs.net
 

ChrisWalker

Active Member
George,

Thank you for the insightful comments. This is a great thread.

Since we spend so much time over in the Z-Wave world, and also spend time talking to SmartLabs, ZigBee Alliance Members, etc., I thought that I'd throw my 2 cents in too...

I have chatted with "ZigBee Alliance" companies, and nobody seems to want to certify on the existing lighting profile (which has been out for a long time now). I think they're waiting to see if some of the core issues (like single points of failure) clear up first. I think they're also waiting for a profile which allows them to interoperate with more than just other lighting devices.

As George mentioned, the "ZigBee" name has become a marketing term rather than an interoperable technology like Bluetooth. A lot of the companies developing and shipping Z-Wave products looked at ZigBee as well, but didn't like the IP issues, standards limitations, absolute lack of any actual "ZigBee" interoperable products, and marketing confusion. I think the whole matter will come back to hurt the ZigBee Alliance in time, which probably isn't good for the industry generally--but it certainly serves as an advantage to them in the short-term, spending megabucks marketing what is still essentially vaporware.

In the end, 802.15.4 is the "TCP" of the ZigBee world--you can make products on it, and they could theoretically talk to each other if they just used a common language. That is what we have today in that world--non-interoperability. And as has been proven by the Internet, it is interoperable standards (like HTTP and HTML) that ultimately win.

Also , one quick clarification:
1. Z-Wave networks can have up to 232 devices. This is for residential applications. Some vendors, like Cooper Lighting, are doing "plug and play" intra-network routing to allow larger networks (since they have warehouses and commercial customers). Basically, you get the same capability of a larger ZigBee network without the extra cost (hardware and development) driven down to end consumers in every last node.

[I know this is off-topic, but has anyone had a chance to play with Wayne-Dalton's new Z-Wave garage door products which turn cars and garage door openers into Z-Wave remotes? Or the new Sirius Conductor Satellite Radio with the Z-Wave metadata remote? They're both awesome.]

Chris
 

Spanky

Senior Member
Zigbee has not caught up with Zwave on the available components available yet. Zigbee is an interesting technology, like Zwave, that merits further attention.

I believe the installer and user interface is the key to success of the technology.
 
Top