Internet of Things (IoT) / Eureka Park @ CES 2015

Eureka Park is one of my favorite things about going to CES. It’s a special place where pretty much anyone can come to the show and demonstrate their product. Most are actual products (like UPPER DESK), some will be soon, and others are still questing for funding. With Internet of Things (IoT) (which BTW is a horrible name, but since it essentially means putting a device on the network and having it report, and potentially change, state, and I think that’s too low a bar to use “Smart Device” or something similar let’s run with it) commanding focus this year, there were a multitude of different network enabled devices on display making it a great thought exercise on IoT innovation.

Before getting to the products I think it’s important to address IoT “fragmentation”. There are too many one-off devices and proprietary protocols right now. Building a comprehensive system off of anything new simply isn’t possible because these things simply can’t communicate with each other in a way that makes them useful beyond the novelty stage (“Wow, I can turn the light on with my phone”) to actually add value to a smart home via automated actions that occur based on event triggers (aka “scenes”), which was a shockingly foreign concept to too many vendors at the show (not just Eureka Park). While the choking levels of incompatibility are frustrating, it is better to allow myriad products to arrive more quickly, in a less compatible state, and worry about interoperability later after natural selection thins the herd.  Most of the vendors I talked to acknowledged that playing-nice-with-others is an important aspect of success, hopefully that was more than lip service. With that out of the way, let’s take a look at some of the more interesting IoT products on display.


The idea behind Roost is simple. Connect existing smoke/CO2 detectors to the network and report their state. How it works is via a special 9V battery that has Wi-Fi and an alarm detector built in. Each device costs $30-$40 (bulk v. individual), the battery should last five years, and when it’s time to replace it you just swap the battery portion (the Wi-Fi/detector module slides off). Right now status (alarmed/low battery) is only reported through the app (via a Cloud hosted service), but because it is IP out of the box it should be relatively straight forward to integrate with other services; something the representative I spoke with acknowledged as a goal.


Connectsense is an $80 Wi-Fi enabled external receptacle with a Zigbee gateway and USB charging built in. Additional Zigbee only receptacles will be available soon to add coverage in the house through the single master controller. Pricing on the Zigbee only version wasn’t available, but if they can keep the price reasonable and provide either a rich home automation controller (HAC) experience (right now it appears to be app driven, and unclear on how much intelligence is possible) or open the API to allow it to tie into a full featured HAC.


The inclusion of environmental sensors (temp, humidity, pollution) in the 260 € e-sylife HAC makes it quite unique in the market. Integrating sensors into the system is a great idea, but I’m not convinced that the approach is sound. The main reason it that for that to be useful the device needs to be out in the open, and while e-sylife acknowledges that by copying the Boxee Box’s styling (wonder if there’s a design patent ;)) even then it will only cover one area, but I don’t want that. Automation kit should be discrete, sensors should blend in, the HAC should live in a corner (or even better a wall) in a room where I don’t have to see or think about it.  


At first glance Switchmate seemed like a great idea (they are still in the funding stage). The Bluetooth LE (BTLE) switch cover attaches over the top of existing [American style] light switches via magnets to make a “dumb” switch “smart”. Using BTLE means that the battery will last 8-12 months and it can leverage an Android or iThingy supporting that protocol. If you want to use it away from home, then the optional hub needs to join the party which makes the switches togglable via IP (although probably still limited to their app). The innovation here is focused solely on the portability and ease of installation, other than that this is just a light switch, er… make that a $50 battery operated light switch, that can’t draw its power from the house. Remembering to charge the battery isn’t the only concern either:

  • is there a significant market of non-permanent “smart” switches?
  • as the number of switches installed in a home grows, so does the maintenance burden
  • state can’t be tracked for three-way switches (i.e. most hallway/entry lights)
  • interaction with a richer ecosystem of devices (door lock, alarm systems, motion sensors, etc.) won’t exist until later and will require purchasing the BTLE->IP hub (price unknown) increasing the installation cost even further