Maintaining the Internet of Things


The past few days, post CES, I’ve been seeing a lot of posts about the challenges of the Internet of Things.

BBH Labs points out that none of these new consumer products work together, and that the result is more burden:

A lack of interoperability across devices and platforms will suck our time, not give it back to us. Endless notifications leave us stuck in an inescapable chain of device control. The traffic is bad. Get the heating to come on later. Delay the slow cooker turning off. Record the show you’ll miss. Get the washing machine to come on later. Stop 3D printing the cake decorations. […] we’re looking at maybe a dozen apps here, all independent and all probably built on the manufacturer’s own proprietary system. If nothing else, the dominance of ‘smart phone controlled devices’ at CES will inevitably mean we all run out of battery about five minutes after leaving the house.

Monday Note explains that while the Internet of Things in the business world is doing well, it’s the consumer space that’s troubled. The post talks about the “basket of remotes” where we end up with so many different remotes (or apps) because none of our devices can communicate with each other, or enable a sort of “control center” that lets us manage things in a unified manner:

Why don’t Consumer Electronics manufacturers provide machine self-description and two-way communication? One possible answer is that they’re engaged in a cost-cutting race to the bottom and thus have no incentive to build more intelligence into their devices.

Ars Technica talks about some of the security issues with putting our household devices online. But also examines how smart devices need up-to-date apps. And considering that software services come-and-go, update their features, and change how we interact with them, it means that our smart devices need to be updated.

Herein lies the problem, because if there’s one thing that companies like Samsung have demonstrated in the past, it’s a total unwillingness to provide a lifetime of software fixes and updates. Even smartphones, which are generally assumed to have a two-year lifecycle (with replacements driven by cheap or “free” contract-subsidized pricing), rarely receive updates for the full two years (Apple’s iPhone being the one notable exception).

It’s especially a problem for long-term items like refrigerators, TVs and cars. And there is little current incentive for their manufacturers to spend money proving support to them. Ars Technica’s worry is that the Internet of Things is bringing planned obsolescence to durable goods:

Replacing an otherwise perfectly good TV ahead of time just because its Netflix app is stale and no longer maintained is a reprehensible waste of resources. I would like to think that most people would recognize the wastefulness this represents and wouldn’t ditch their TV just because its built-in Netflix app is out of date. But I’m confident that such thoughts have entered the minds of TV company executives, and they’re hoping people do precisely that.

BBH Labs is a bit more optimistic. Hoping for a solution to this problem:

So, for 2014 and then ahead to CES 2015, I’m less interested in the devices themselves, but instead the platforms and systems that bring them together. Will we see an open platform and data standards for device control and tracking, allowing developers to add the cross device connectedness that the manufacturers can’t?

All of these ring true to me from my own experiences. Having things like Philips Hue lightbulbs and a Sonos music system are pretty cool. The give me control and options that I didn’t easily have before. But my nagging worry has been that these products will last only as long as their manufacturers update and enhance the apps to control them. If Philips doesn’t update their iPhone app, then at some point I’m going to be stuck with lightbulbs I can’t control any more. And, with a product that has an expected lifespan of 25 years, and given the thoughts above, that seems like a real possibility.

Time for us to all roll our sleeves up and work on a solution.

I’d also add, as a footnote, that having worked for years at BMW, a company which has very long lead times for product development, developing products (ie. cars) for technology that isn’t out yet is really hard. Automotive companies struggle between building the technology into the car — with interfaces that are optimized for driving conditions, and that are “on-brand” — and allowing customers to bring their own technology and let it have its own lifecycle. But I’ll save that for another post…


Since writing this post I’ve come across a couple other posts…

The first, from Frog’s Design Mind, talks about the value of the data that can be generated from connected devices. Their breakdown into immediate, aggregate and latent values is quite interesting. They put out a good call for designers to “provide a human lens to better understand, interpret and pinpoint value from the data we are reaping. And I like the comment that as we think about connected “things” we can reinvent what those things are:

…not everything should be connected. Sensors and connectivity don’t always result in a better experience, increased sales or usable information. […] Equally important, let’s not be afraid to make something new, because we can’t rely on the notion that old things will be inherently better once they are connected.

Second, Wolfram’s Connected Devices Project seems like it’s aiming to solve the infrastructure problem that BBH Labs referenced. Wolfram’s got a lot of data about a lot of different devices, and from their About page, it looks like they’ve got a language and way to provide drivers for working with the devices.

Lastly (for now), is this article from Wired, on how wildly insecure the Internet of Things is, explains the reasons it’s so hard for manufacturers to keep the software in their products up-to-date. Specifically, the article details how routers are made by a stack of third party manufacturers and software providers. The result is that the final brand on the product doesn’t control much of the underlying software — instead just adding some new features and an updated UI. So the product has 4-5 year old software, and often no one has the source code any more.

The result is hundreds of millions of devices that have been sitting on the Internet, unpatched and insecure, for the last five to ten years. […] And the Internet of Things will only make this problem worse, as the Internet — as well as our homes and bodies — becomes flooded with new embedded devices that will be equally poorly maintained and unpatchable.

(Image source: link)

UPDATE: I’ve published an updated and revised version of this post on Medium: