CES 2017 UX of IoT report

This is a personal and targeted report on CES. I didn’t look at any TVs, phones, cars, or refrigerators with cameras. I’m not looking for any killer new consumer appliances. I go with my “Beyond Mobile” approach which has a very specific set of goals in mind:

  • See what’s happening in home automation
  • Check out any new wireless technologies
  • Look for any clever UX technologies or new approaches

Here is a list of what I’ve found in no particular order.

Hubs, Hubs, Hubs
In talking to a few IoT pals it was clear that there is little consolidation happening at the wireless levels. There is a sad resignation that we’ll have hubs for some time to come. There are multiple wireless tribes: BLE, Z-Wave, ZigBee, and if you can believe it, a new one called ULE, each just forging full steam ahead. I must have found over 20 booths selling the nearly exact set of sensors and actuators that all work with their specific hub. In discussing with their floor reps, it’s clear they REALLY want you to use their devices. If they belong to one of the above tribes, they’ll let you use other devices from their broader tribe but there was a clear pause and an implicit warning. This implies the overall compatibility, even within these tribes, is still a moving target.

As a fairly technical guy I’m very reluctant to enter this space. I can’t imagine making a big investment in my house right now. Who wants to update 40 light switches/outlets in their home, at $60 a pop, only to find out that they’ll be incompatible next year when the tech winds shift ever so slightly?

Retro IoT concepts
What surprised me was a super simple edge product called SwitchFlip that was clearly a reaction to this complexity. It is nothing more that a set of paired outlet plugs. You plug device A into a switched outlet A and device B into a normal outlet B. When you flip the switch to turn on outlet A, outlet B turns on. Very simple with no hubs, no programming, and no hassle. It’s not for me but the amateur ethnographer in me finds this fascinating. Entire companies are forming around the frustration of the current thinking around the smart home ‘wireless hub’ concept.

The UX of automation
Speaking of these hubs, so many have the same tired approach: add a device, create a room, hook up devices to events. The entire process is designed around turning on home lighting with motion sensors. Now, new products like Vivint’s Sky did announce a simple home AI that can do things like detect an empty house and power things down as well as close your garage door. This is a very interesting direction and represents more what I expect people will want in this space: devices that organize themselves and do things without me pulling out my app. We have to explore the value beyond the current complex-setup-and-fixed-rule-based-triggering. I find this assistant direction offers users much more value and less hassle.

In fact, my big UX takeaway from CES was that there was no killer screen/device/haptic technology that everyone was talking about. For me the big UX story is how it is moving behind the scenes for more automated AI approaches (e.g. NVidia, Toyota’s Yui). The really cool demos involved the user doing nothing at all. The house is just saving you energy, the car analyzing the road ahead, or the shoes recording your run automatically. The best UI is no UI and it showed at CES.

Would you calm down about Alexa?
If you read anything about CES you’ll find many folks going on and on how Alexa ‘won’ CES. Yes, it was clear that lots of companies are integrating with it. But in my very personal opinion, it’s because Alexa’s API was launched much earlier than others. In addition, it’s not really clear what this adds to many of these products. There is no doubt that voice interfaces offer some clear value which will, of course, grow over time. But there seems to be so many bigger fish to fry (like discovering the right value to these devices in the first place). I’m not anti-voice interface, it just feels like a local maximum to me.

The up and coming (and long term) BLE Mesh standard
I had some exciting talks with a few companies that currently support a proprietary BLE mesh solution (there are at least 4 BLE mesh implementations) but they enthusiastically support the new BLE Mesh standard once it finalizes. This is very heartening and gives me hope that some day, we’ll all use the same low level standard. However, we’re at least a year away, if not more. The advantages of the BLE Mesh standard are huge. The biggest in my mind is the death of the hub. To be sure, there is value in having one, but it isn’t required and that’s a big step. In addition, all commands are local, with no need to go through the cloud. Finally, as a tinkerer, I can add my device to pull some devices strings but it’s just another device, it’s not a central, required hub.

Moving beyond sensors
It’s astounding how fast hardware is moving. There were many unknown Chinese companies showing off very impressive smart device concepts. Lots of very cool IoT platform and prototyping systems like Bread and SimBLE as well as ‘introduction to China’ companies like HWTrek that are making it easier than ever to build and test devices. The hardware roadmap for tiny, energy efficient yet still very capable processors (like the Nordic nRF52) implies that we’ll be able to do quite a lot of computation at the edges with these devices in the very near future. In fact, I have a theory that much of cloud computing will just move to the edges as we get more and more capable edge devices. (note: Lee Ratliff politely pointed out that this is called “fog computing”) It seems likely that hardware capabilities will be the big driver in new product concepts (assuming we can securely communicate of course, but that’s another reason I’m bullish on BLE Mesh)

I find CES to be a very hard conference to read. There are so many things happening that getting a clear theme is illusive. My primary take away is that companies really really REALLY want to make home automation systems but it still just feels too proprietary, and therefore risky, for most consumers to really buy into it. There are some very exciting things happening but this is going to be a game of years, not months.