A Bluetooth scale! A Wi-Fi door lock! And, of course, a fridge that tweets! We can joke all we want, but smart objects are here to stay. And if anything, we’ll only see more of them. But everyday appliances have a funny problem: they are getting smarter (read: connected) and dumber at the same time. UX people: it’s high time you fixed that, unless you want to earn your place in history at the receiving end of a tweet from @internetofshit.
Last year a friend told me about a founder touring Silicon Valley to raise funds for an Internet of Things startup. The founder was building a solution for home appliance manufacturers interested in connected devices. Smart kettles. Smart lights. Smart everything. One investor allegedly interrupted the pitch. “Do you know how I usually turn on a kettle? Like this” he said as he flipped the switch on an electric kettle. Suffice to say, his checkbook remained closed.
This is a design problem. You might replace the analog switch with an app, claim it’s the future, and proudly exhibit at CES. But in fact you have un-designed a perfectly good kettle.
Connectivity opens up so many possibilities it’s hard to capitalize on them and retain the simplicity of an analog interface with two buttons and a knob. Just think about the mobile landscape in the pre-iPhone days. Smartphones appeared on the market with tons of features but the mobile web was limited, navigation confusing, and performance painfully slow. I still remember my Nokia E52 (high-end handset in its day!). It notoriously froze when typing, which rendered texting borderline impossible and trauma-inducing. It took Apple’s touch of brilliance to actually put smart in smartphone.
With smart objects, we’re still in the Nokia E52 days.
Fun fact: E52 was introduced after the iPhone, but UX-wise it was so pre-iPhone, we’ll stick with it as an example.
Quest For Usability
Design problems with many smart objects start right after unboxing. Analog devices have often been trivial to set up. Even with bad design there’s only so much trouble changing a lightbulb can cause. Unless it’s a smart one. Hue requires you to connect lights to a bridge – a piece of hardware you must hook up with your Wi-Fi router, while with LIFX you’ll need to sign up inside their app, connect to another network, and then add a bulb to this network. And it’s not just bulbs. Watch this setup video for Fitbit Aria. When a $100+ bathroom scale asks you to manipulate its batteries to enter setup mode, you know something went wrong.
The problems don’t end with setup. Take Lockitron for example: is it really faster to open the door with an app that has to first connect to the lock over Bluetooth? So maybe it’s an issue with devices that need companion apps, because they don’t have a native GUI? You wish. Smart TVs jumped on the connected objects bandwagon years ago with GUIs spread over 40-inch displays and still are notorious for buggy interfaces. And so are most smart home devices. Experience offered by most smart devices isn’t worth the struggle even for geeks. And with the rising popularity of connected appliances, manufacturers need to start designing them for the average Joe. No one sums it up better than Fortune’s Stacey Higginbotham:
The smart home has a problem. While people are interested in the technology, they also aren’t ready to buy it. And I don’t blame them. A lack of standards, a disregard for usability, and an incoherent story about what a smart home can do for people all mean that anyone interested in buying a connected product quickly encounters a cautionary tale that makes them think twice about spending $200 on a connected door lock.
The Promise Of Language
Good news? We are approaching a new dawn for smart objects thanks to Amazon and their Echo speaker/virtual assistant, already touted as the iPhone of the smart home. While everyone in the connected devices market was busy designing a companion app that would not be awful, Amazon ditched the GUI as an inferior approach to interaction with objects in the physical world. They replaced it with voice control and people love it. That’s because language feels natural: something that cannot be said about a two-dimensional graphical interface.
The advances being made in speech recognition, natural language processing, and deep learning bring us closer to the moment where every vendor will be able to create a language-based interface with off-the-shelf components (in English, at least). Even today it’s not a stretch: Amazon exposes an API for Echo’s underlying AI, Alexa. Thanks to that, third party devices and services can already use Echo as the interface. And it’s not only about remote control. Amazon imagines Alexa as a platform. A gateway to a number of devices and services, not far off from how iPhone became a gateway to apps.
Good as it sounds, the success of Echo won’t suddenly make smart objects actually smart. There’s still a ton of work ahead of UX designers. Echo introduces a new paradigm, just like iOS did in 2007. But even on iOS, with system-level limitations, a host of available frameworks, and Apple’s design principles enforced by the App Store reviewers, there’s more shitty apps than quality ones. UX designers now have to establish best practices. They have to learn – and teach the users – about language-based interaction. And they have to figure out frictionless onboarding: Google’s Scott Jenson is right when he points out that smart home really needs an automatic configuration standard.
The arrival of iPhone has erased the previous concept of a smartphone from our collective memory (and my experience with Nokia E52 tells me it’s a good thing). Amazon Echo can do the same for smart devices. UX and interaction specialists, however, need to pull themselves together and use the groundwork laid by Amazon to finally create seamless experiences for the connected objects. Otherwise, I’m sticking with an analog kettle.