Google Enters the HA Market – Android@Home
Since writing this entry, little more information has trickled out:
- The LED lighting OEM appears to be company called 'Lighting Science' who plan to offer LED light units at around $30 price by end of 2011.
- The wireless sensor network protocol appears to be based on 802.15.4 and 6LoWPAN, another piece of good news
Google made an announcement this week at Google I/O conference where at the keynote they demonstrated an Android device controlling the lights of the showroom. This created quite a bit of buzz in the HA field, and we've since received quite a few questions what exactly does Google's move mean?
We've tried to figure it out but things are quite fuzzy still, and lots of detail is missing.
First up was the intent : the big announcements were around media streaming service (Amazon very recently announced the same) and the Android Accessory API (based on Arduino, yay!). That makes sense to Google from strategic point of view and both are directly aimed at Apple.
In that context the whole HA announcement seemed a little bit of a side-show. So we were trying to figure out if Google the advertising company is getting into HA with a real push (serious partnership money required from their part) or whether the demo was more of a "party-trick" to show what could be done with the Android Accessories vs. what Google will put serious weight behind in terms of creating a new market opportunity for them - and for the record, getting a standardized USB API from Android side to dedicated HA extensions would have tons of potential.
The one tidbit that was revealed there was that they are working with some LED lights OEMs to demo light switches by end of 2011. That is the interesting bit to watch. If they do plan on building a serious HA ecosystem around Android, establishing similar relationships they've done with Samsung, HTC, Motorola, Sony Ericsson in the cell phone world would be necessary. Google should be looking for partners in the area of HVAC, Security, multi-media (they've got Google TV, that's a good start), even healthcare to get to a complete residential automation solution. It's a large, extremely fragmented field and a big name like Google pushing for some unification would be very welcome.
So watching whether there's a desire, will (investment) and understanding of the need to build the partnership network is one way of measuring whether this effort will take off or if it will fizzle like a Google Wave.
The second part is the wireless protocol itself that was mentioned. Sounded like could be a new proprietary entrant to the wireless sensor network field because little detail was given. Whether it builds upon 802.15.4, 6LowPAN or anything in that direction would be interesting to know. It can potentially make things more complicated in terms of technology choices if it is a new entry, or it can put some serious weight behind one of the existing WSN options if one has been chosen. Will be very interesting to see what emerges there too.
Getting one clear direction on wireless sensor transport would be very welcome (we like 6LowPAN here).
Whatever the case, and especially if its a new wireless protocol, the approach Google seems to be taking there is to get a vendor buy-in (the previous point about a LED light OEM) - and for a company with the money and brand recognition of Google that does make sense. It does not however remove the reality of how slowly consumer space updates hardware in general (do you want to buy a new TV again?) and residential buildings in particular (your boiler is from what year again?). So in those terms if we think about the HA market as a whole where a fully automated solutions are possible, integration will still be key for long time to come. No details yet on how legacy integrates with the all-new Google HA with from-scratch approach. Yet another key point to watch as the year progresses.
And having said that, if the Google effort will be successful and there's a serious push on this, getting something to emerge long-term as potentially ubiquitous HA stack would be a very big deal.
The other great long-term side-effect of ubiquitous runtime (Android) and protocol (currently unknown) is that it would completely commoditize the controller and transport protocols which would benefit everyone and grow the market place immensely. Focus would then move to building tools for professionals and pro-sumers.
What Does It Mean For OpenRemote
So we were thinking what would an announcement like the above mean for OpenRemote. Mind you there are big caveats here, first with if Google is really serious about HA or if they just wanted to build a party-trick to make Accessory API look cool and second if they've analyzed what it would really mean for them to enter a HA market (they are an advertizing company after all).
Anyway, let's assume Google is serious about it because that is the most exciting scenario
In terms of controller, the runtime would clearly shift towards Android. We've been going back and forth about this a for long time, not really getting to a decision whether go with a full Java SE or go with Android (great for embedded which is where the controller will eventually live).
On a tangent, I'm gonna take a dig at Google here for messing up the Java compatibility issue with Android. They dropped the ball there. Had they stuck with WORA the above would be a non-decision. Now Android is an extra effort. Boo to Google for not living up to Java's WORA promise. Double-boo.
Ok, that out of the way, controller running on Android would become a clear target. The big extra effort there is to port all the cool Java libraries (rule engines, scripting languages, and the such) that we want to use to Android. But once done, an Android controller is easily embedded into very small devices, merged with your panel hardware, hidden away, whatever. One less piece of hardware and one less install to deal with. Goodness.
In terms of the wireless protocol that we know nothing about, it would be an obvious candidate for a back-bone transport protocol that is much, much needed. We've got the IP Ethernet for wired installations but we do need a clear back-bone for the wireless sensors (and WiFi isn't it). At OpenRemote we don't care if that ends up being Z-Wave, Zigbee, 6LoWPan, or something from Google (but we like 6LoWPAN, did I say?) because we are agnostic but having some clarity in the industry of how frames are transported between controllers, sensors and devices would certainly be a big deal. For everyone. More time spent on building what matters and less time spent on twiddling bits trying to map signed and unsigned, big-endian and little-endian, whatever.
Finally, the Android Accessory DevKit will be very very interesting - and the only part that looks more or less concrete right now. Whether its an IR extender to your Android Hardware device, or a wall-mount frame (maybe with some key physical buttons) for your Android tablet it will create a standard eco-system of hardware extensions. This wasn't targeted by Google to HA specifically but it will make a difference in HA and will be a big boon for anybody wanting to build Android-based and Android-integrated HA extensions.
This Will Be Good
Bottom-line: exciting announcement from Google!
- Hope they're serious about HA beyond producing a cute party-trick at keynote.
- Moving towards a less-fragmented ecosystem with Android as runtime, Android Accessory API and wireless transport would be a huge deal for everybody
- Watching and waiting for some serious partner network push here
- Using Arduino is awesome!
- Getting some installers pulled into this because of Google brand and weight would sure help
- OpenRemote will for sure want to participate
- At the very minimum the Android Accessory API and using Arduino will help.
What do you think? Please share your thoughts on this forum thread that was started.