Google IoT OS 'Brillo' Effort Reports to Barratt, Targeting Home Routers

Yesterday, The Information reported on a new effort by Google to create new software for the Internet of Things. The Information reported it would run under the Android brand and would connect low power devices such as fridges, light bulbs, garden monitors and more. Brillo likely would be part of the broader Android family.

I asked around from my sources and got some more information on Brillo not found anywhere else. This is what I learned:

Google’s been out talking to manufacturers about Brillo for a few months. One of the key early targets for Google was home router manufacturers. This is because as part of the local “smart things" network architecture envisioned by Google (at least in the home), there would be a set of code (likely a full RTOS) running on a home router that would enable it to talk to low power devices and act essentially as the smart home hub. According to my sources, Google has approached numerous router manufacturers such as Netgear, Asus and possibly others. They say Google sees existing home device categories - particularly the home router -as a very logical device to act as a central part of the smart home architecture. It’s essentially a post-hub vision (a direction for the smart home I’ve been writing about for over a year). Basically Google wants Brillo to do for routers what Android did for phones.

Much of the compute will run in the cloud. My sources compared it to the same overall concept as Google Chrome OS, which as many know was an architecture that utilized an ultra thin software client OS for Google Chromebooks and runs most of the compute in the cloud.

With Brillo, my sources say Google is talking about utilizing low power radio technologies other than Thread (a Nest-initiated effort that's now an industry consortium) such as Zigbee. The reason is Google knows they would need to talk to a variety of installed legacy nodes currently running out there today, many of which have Zigbee radios that wouldn’t necessarily be field upgradable to Thread (both technologies run atop the IEEE 802.15.4). My sources say they're also talking Thread as well, but the news here to me is that it's bigger and beyond Thread.

The team working on Brillo, according to my sources, reports to Craig Barratt, who many will recognize as the former CEO of Atheros (which later became Qualcomm Atheros after Qualcomm acquired the Wi-Fi chipmaker). Barratt, as you can see from his Linkedin, has been a VP with Google since June 2013. Brillo is also very much not part of the hardware group being run by Tony Fadell, as Barratt’s overall responsibility includes the Google access efforts such as Google Fiber, Loon, Drones.

All of this makes sense to me. A few thoughts based on what I’ve learned:

  • The Information stated Brillo would run on low power devices with 32 to 64 MB of RAM. The problem here is that much RAM is not typical of most ultra low power devices in the home. Smart locks or smart bulbs have much less than that, typically running something like 128KB of internal RAM, an order of magnitude lower than 64 MB or RAM. This validates the router-centric network architecture from my sources. While I can envision other devices other than a home router running the full Brillo stack (like maybe, say, a Nest Thermostat?) and bridging to these truly low-power, low-memory devices, the home router has always been the most logical device category to absorb the home IoT hub capabilities. 
  • The vision allows Google to insert itself into the smart home and consumer IoT use cases without relying on their own native hardware. While Nest and Dropcam are doing fairly well in a nascent smart home market, the reality is there are a variety of platforms out there and Nest is throwing elbows with the likes of Apple, Amazon and likely Microsoft in the future. This will allow them to align themselves with the popular brands already running the biggest network in the home today, the existing home network. 
  • Since the company has been talking about radio technologies other than Wi-Fi and targeting home router makers, I expect they are talking to these companies about building home routers with new radios. This is a new model, obviously, except for some very forward leaning home router makers such as Securifi (who started shipping a home router with embedded Zigbee/Z-Wave this year). I imagine Google’s probably been talking to Securifi as well (or at least they should be).
  • Given this is not part of Nest, does this make sense to have essentially what is a competing vision for consumer IoT and smart home? Maybe, but it certainly fits with past Google behavior. Google’s had a long history of competing and overlapping efforts; witness its multiple efforts to reach us with consumer video with Android TV, Chromecast, YouTube, etc etc. 

That’s it. I’ll have more thoughts next week on this as we learn more. Keep an eye out for the full details at Google I/O next week.

Want to receive marketing intelligence like this in your inbox every week? Subscribe to the Smart Home Weekly. Click here.


  • Reza Hashemi

    Integrating all the wireless HAN protocols at one device is meaningful. That makes home router manufacturers to build WiFi (WLAN) routers that are smart home coordinators too, support for LP-WPAN (IEEE 802.15.4/ZigBee/Thread).

  • Jian Huang

    It looks like that the IOT home gateway approach needs a radio technology that has to have the long multi-room range as Wi-Fi and the little consumption power as IEEE 802.15.4 or BLE ;-)

  • Adam F.

    Homekit isn’t going to be the solution since it’s not open. Bill’s completely right. By limiting it to half the world of devices, that’s an inherent problem. If Google can pull this off it’s pretty huge.

  • Bill Sheppard

    Given that HomeKit is not only iOS-centric, but iOS exclusive, it makes sense for Google to offer an alternate ecosystem which embraces rather than excludes. If they keep it open and provide vendors the flexibility to manage their own data and support varying architectures I think we could well see Brillo with far broader adoption than HomeKit.

  • Jason

    Great scoop. This makes total sense. The router should be the smart home hub. Interesting it’s completely separate from Nest.

Leave a comment

Please note, comments must be approved before they are published