To Automate Everything, Solve These Three Challenges
Just as the internet of things is poised to remake our homes and offices, it’s facing perhaps its most critical test: adoption by the average consumer.
“It was a pain,” my dad says of the programmable thermostat we had in my childhood home. It was a plastic rectangle the size of a small envelope that jutted out from the wall. He would flip down the protective cover and tap the buttons to make the numbers on the digital display slowly tick up and down.
“I could never remember how to do it. I would have to find the manual and try to read and do it at the same time,” he recalls. “When the power went out, we would have to start over.” His countless minutes of frustrated button-pushing made sure the house was cool during muggy summers or toasty when we woke up on cold winter mornings—a common occurrence in Minnesota. Yet, despite his best efforts, there were times when the temperature still wasn’t quite right. The air conditioner would keep blowing, for example, when we left the house on weekends. And he’d have to manually override the settings to keep out the chill of a blustery snow day.
It’s hard to imagine that same thermostat being cutting edge, but that’s exactly what it was more than 20 years ago. It was programmable, but in no way was it intelligent. The only intelligence it contained was that which the user transferred to it. Today, most people’s thermostats aren’t much different, but that’s starting to change.
Thermostats are relatively banal devices, but as the startup Nest has shown, they are ripe for an overhaul. Nest’s thermostats can sense when people are home, monitor weather reports, and respond to commands from a smartphone, adjusting the temperature as needed. “In some ways, it is a first step in the right direction,” says Lorna Goulden, a technology consultant, regarding Nest and the incoming tide of smart devices. Indeed, the thermostat is just one of a slew of new devices that blur the boundary between virtual and physical worlds. They promise to make our built environments more intelligent, responsive, and efficient.
Collectively, they are called the internet of things or the internet of everything. The concept was first envisioned back in the late 1990s during the dot-com bubble. Then, people anticipated a near-future where computers and the internet were everywhere. Now, fourteen years later, connected devices are just beginning to slip into homes, from smart thermostats to apps that unlock your door without a key. Smart objects are slowly transitioning from pioneering to practical.
Yet just as the internet of things is poised to remake our homes and offices, it’s facing perhaps its most critical test: adoption by the average consumer. The intelligent future promised by entrepreneurs won’t catch on if those devices can’t connect to each other automatically, lack intuitive programmability, or aren’t appealing designed. If they fail at any one of these, automating our homes may be more trouble than its worth.
But if engineers and designers can nail each of those requirements, then much like electricity did a century ago, the internet will course through our homes so seamlessly we may pay little attention to it as we go about our daily lives.
Connecting Everything
There’s a good chance that you already have an internet-connected device in your home. It may be a DVR or a set of wireless speakers. These are early components of the internet of things, but they’re missing something important—the ability to interact with the other objects in your home. “They’ve got a little bit of a learning curve,” says Craig Miller, vice president of worldwide marketing at Sequans Communications, which creates computer chips for smart devices.
The most intelligent smart devices may need not just one connection, but two: one to the internet and another to fellow smart objects. We know how to handle the first one—just add a chip for Wi-Fi or cellular service like LTE and you’re online. Coordinating with other smart objects can pose a problem, though. Currently, device-to-device communication is experiencing some growing pains, much like wireless networking for computers did in the 1990s. It took a few years for competitors to settle on a standard, Wi-Fi, and a few more before it became widely adopted.
Users don’t want to—and shouldn’t have to—think about how their refrigerator might talk with their dishwasher.
“It’s not a lack of standards, it’s the fire hose of standards that’s the problem,” says Rob Faludi, an adjunct professor in the Interactive Telecommunications program at New York University and chief innovator at Digi International, a networking company. “But that’s always a problem with devices,” he adds, referring to the historical differences between computer platforms such as Windows and Mac. In that case, users felt computers were valuable enough to put up with the problem of incompatible devices. Eventually, a consensus developed around key standards, and the market coalesced around them.
For now, the torrent of different networking standards poses a problem for average users, who don’t want to—and shouldn’t have to—think about how their refrigerator might talk with their dishwasher. Most smart devices currently require users to delve into application programming interfaces, known as APIs, for which you need a good deal of programming knowledge to use properly. Companies such as Microsoft are hoping to simplify the operation of a smart home, building dashboards that allow users to control disparate smart objects from their PCs or an all-controlling smartphone. But even these tools require some effort and knowledge to implement. Simplifying this process won’t be easy—the more devices in a network, the harder it is to coordinate.
Engineers have a long road ahead of them, but if they can make communication seamless, there’s a lot of potential in that connectivity. Faludi offers the example of a home air conditioner and a security system: Say it’s been a hot three days and the air conditioner hasn’t turned on. There’s a good chance the occupant isn’t home, and, if the front door is unlocked, it might be time to close that deadbolt.
Programming It All
Connecting devices is only part of the challenge. To really unlock the power of the internet of things, smart devices’ functions must be accessible to average users without making them cede too much power. If people don’t feel in control, they’ll be hesitant to adopt the technology. “It can be very discomforting to come into a house and all [these] things start happening,” explains Jason Johnson, co-founder of the Internet of Things Consortium. In reality, intelligent homes will be only partly automated, giving users final say over what happens, just like our current relationship with computers.
What people really want is “the Apple experience.”
Even one of the closest examples of full automation today, Alex Hawkinson’s SmartThings, prompts users for guidance at the outset and later allows its decisions to be overridden through a smartphone app. For example, objects connected to the bathroom fan and the shower faucet may prompt the user whether they want the vent to turn on or the heater to fire up. Another may connect to the blinds and default to raise in the morning and lower in the evening, but will still allow people to raise and lower them manually. That way, there aren’t any surprises.
As more smart devices are added to a home, the number of possibilities—but also the complexity—could grow exponentially. Finding the balance will be tricky. “If the usability is lacking at any step of the way, and frequent updates present more frustration than excitement and delight,” explains Goulden, the consultant, “then interest will quickly fade.” What people really want, Goulden says, is “the Apple experience”—pull it out of the box and it’s ready to start using.
Designing for Everyone
There’s one final, and often overlooked, challenge—design. It’s easy to see how bad design can frustrate a user—take the baffling array of buttons on the thermostat at my parents’ home. The buttons’ position, hidden behind a panel, made the whole package nicer to look at but not any easier to use.
Contrast that with Nest, which has no buttons. Rather, its physical interface consists of one rotary dial. Not only does it pay homage to old thermostats—making the device seem less threatening—it’s also intuitive: turn right to raise the temperature, left to lower it. “It was born out of frustration, which I think many people can relate to when they’ve tried to program their programmable thermostat,” says Kate Brinks, director of communications for Nest.
The best, easiest-to-use smart objects will likely look no different than devices we use today, Faludi points out. “A big chunk of this will just be baked into things that we buy,” he says. “You won’t buy an ‘internet of things.’ ”
There will be entirely new products, but that shouldn’t untether design from reality, Johnson says. “Technology products shouldn’t look like technology products.” Developers should shoot for “something that either spouse could bring home and put on the kitchen counter,” he says. “And it gets past that ‘Ew, what is that thing.’ ”
And rather than just being “smart” for technology’s sake, the devices should also address people’s actual needs, he says. “We need to develop products that are very practical, that are solving very real problems,” he says. “Not just, ‘Gee whiz, wouldn’t it be cool if I could turn on the lights from my phone?’ ”
Goulden adds that, in her consulting, she advises clients that smart objects should act as an extension of the user. “How do you take an individual’s identity and how do you relate that to the objects that are around them?” The more relatable the object, the simpler it is to understand.
Too Smart a Future?
As the internet of things becomes a larger—and less visible—part of our lives, it could change the entire meaning of privacy. The media went wild in July when two security researchers turned a Jeep Cherokee into a child’s toy, controlling it remotely through the car’s digital diagnostic port. But future security hacks need not be so flashy to be concerning. The internet of things has the potential make not just individuals but hospitals, governments, and cities vulnerable. If everything—from medical devices like pacemakers or pill dispensers to infrastructure like bridges and railroads—is on the network, then the consequences can be deadly.
The number of vulnerable points—the “surface area” for attacks—with a connected device is often greater than people realize, says Chris Poulin, a research strategist with IBM Security. For example, if you have a Nest, it’s not just that someone could get into the physical blob on your wall. There’s also, “the mobile app, which connects to Nest’s [data servers], and then to the Nest in the house, which connects to the WiFi, which again connects to Nest’s data servers…the surface is just really broad.”
That’s part of what the Federal Trade Commission worried about in their report last January . People need to consider not just unauthorized access to their smart fridge or connected garage door, they stated. Consumers should also be aware that their personal information being collected from their smart objects could potentially be accessed by future banks, employers, or insurers who might make decisions based on that information.
In September, the FBI released a cyber crime public service announcement , explaining how the internet of things can be exploited. Among their recommendations for how to defend against those exploits is a simple suggestion: “Consider whether IoT devices are ideal for their intended purpose.” In other words, the first step for a consumer should be, is that smart fridge really necessary? Lee Tien, a specialist in privacy and civil liberties at the Electronic Frontier Foundation, says consumers must watch out for themselves because the companies aren’t always able to. Especially when devices are made by small startups, he says, and “not in the first tier of expertise of security.” Bruce Schneier, a computer security and privacy expert, says after devices are shipped, companies have no incentive to update old software or computer chips to ensure future loopholes are fixed—they’re too busy working on their next thing. He says until the makers of devices go open-source and allow outside engineers to help find and fix security problems, this will always be the case. However, the new problems are slowly growing interesting solutions. Companies like Honeywell that once were known for protecting your home now protect industrial plants working in oil, gas, chemicals, or minerals and use the internet of things. For consumers, Schneier says large companies that control your entire ecosystem of technology, whether it be Google or anyone else, will be how people keep their hardware and software up-to-date. “Apple does security for all iOS users. This is the future.” Yet there is a contingent that doesn’t see privacy concerns holding up adoption. Faludi argues that people are “always going to have concerns about security and privacy,” adding that, “they’re problems for developers to solve, but they’re not really any kind of barrier.” Researchers like Faludi are wary that a lack of confidence in the future could stop adoption.
On the Cusp
Despite the hurdles, the potential for the internet of things is enormous. Not only could it simplify many aspects of our daily lives, it could also make our homes more energy efficient, saving us money and reducing our environmental footprint. And few stand to benefit more than people with disabilities. For some, even simple tasks take an inordinate amount of time and effort. Automating those would allow them to direct their energy toward more important things in their lives.
It may take some time for that to happen, though. Smart objects are starting to trickle into the marketplace, but widespread use of the internet of things is still five to ten years off, according to industry analysts at Gartner, a market research firm. That’s good news for the engineers, developers, and designers who are trying to work out the many kinks that remain.
Should they succeed in making smart objects intuitive, transparent, and minimally invasive, though it’s likely that people like my parents, who are sick of tapping buttons to program everyday life, will adopt them. There are signs they’re getting closer; My dad called the other day, clearly smiling through the phone. “Your mom and I are thinking about getting a Nest.”
This post was originally published on October 16, 2013. The section “Too Smart a Future?” was updated on October 14, 2015.
Photo credits: Nest Labs, philips_newscenter/Flickr (CC-BY-NC-ND)