Close Mobile Menu

The Case of the Hacked Refrigerator—Could “The Internet of Things” Connect Everything?

February 5, 2014
by Glen Martin

Somebody hacked a refrigerator recently, and it could mark a tipping point for civilization. 

It’s no joke (although it is pretty funny). The target was a “smart” refrigerator:  a software-enhanced appliance capable of linking to the internet and sending and receiving information. Why would your refrigerator need to be smart?

Well, it has to do with “wanting” more than “needing,” of course. Say you’re the refrigerator’s owner. Maybe you’re vacationing outside Ouagadougou, and want to ensure the bottle of Cristal on the fridge’s lower shelf is sufficiently chilled for your return. (Drinking a slightly tepid bubbly simply wouldn’t do after a sojourn in the desert fastness of Burkina Faso). So you get on your iPhone and text your refrigerator to crank it down a few degrees.

The problem is that if you can chat with your refrigerator, so can any reasonably adroit hacker. As revealed by the security firm Proofpoint, the recent fridge attack was part of a broader hack offensive against a variety of connected gadgets that included televisions and home multi-media centers. The intent was augmentation of the hacker’s “botnet”—a legion of compromised devices used to overwhelm web sites with malicious cyber-traffic. Proofpoint did not name the manufacturer of the refrigerator or other products that hackers used to dispatch 750,000 malicious emails.

But the violated smart fridge points to a bigger issue: the imminence of the Internet of Things. We all know what the internet is—the global linking of computers and smart mobile devices via cable and wireless connections. The Internet of Things kicks this conceit up a very large notch, implying as it does the connection of, yeah, everything: refrigerators, building heating and cooling systems, cars, coffee pots, sewage treatment and power plants, oil refineries, even implanted medical devices. (The latter being the subject of a recent Homeland episode, where the Vice President’s hacked pacemaker launched him into tachycardic oblivion.)

Add to that the myriad cheap sensors now used to monitor everything from traffic to air pollution to, well, maybe your bedroom window. All this stuff will be—is—generating gigachunks of data, information that can be tweaked and teased and collated and employed for good and ill.

The positive view: A world where everything is hooked to each and every other thing will allow efficiencies of exquisite degree. When energy demand fluctuates, power plant output can respond in milliseconds. No need to guess when the next bus will arrive at you corner stop; an app on your mobile device will plot each and every bus on the metro system. Sensors will allow you and your physician to track your blood chemistry and digestion in real time.

But some of the downsides are also clear, and most involve privacy. A thoroughly connected world connects you to everything and everybody else. At any given moment, the world—if it is nosy—could know where you and what you are doing, and maybe your triglycerides and blood glucose numbers to boot.  And it’s just too bad if you don’t like it.

“Obviously, when every device you interact with is connected, you don’t have much opportunity to opt out,” observes Michael Franklin, professor of computer science at UC Berkeley. “I’m a technologist and an optimist, so I tend to focus on the dramatic possibilities for efficiency it will bring. Am I concerned about security? Yes. But I feel there will be solutions, and they’re probably going to be driven by policy and law. I don’t think we’ll stop the collection of data, but we’ll refine who gets access to it. There will be a lot of court cases, and lots of legislative activity.”

Indeed, the Internet of Things will likely expand the data-wrangling sector by an unprecedented degree.

The Internet of Things has so much inherent momentum that it really doesn’t matter what we think.

“I think the sheer volume of data that’ll be generated could create an entirely new industry,” Franklin says. “We’re going to have to rethink how we collect, process and manage this new volume of information, how we distribute it geographically. Also, how we power it. It will take a great deal of electricity, and we’ll need to develop more sustainable sources and better storage systems.”

In a way, parsing the “good” and “bad” impacts of the Internet of Things is beside the point. It is just one of those developments—like the discovery of fire, movable type, the steam engine, or the “first” internet, for that matter—with so much inherent momentum that it really doesn’t matter what we think.

“Well, yes, it’s a manifestation of the natural course of electronic evolution,” says David Culler, chairman of Berkeley’s Department of Electrical Engineering and Computer Sciences. “It’s not that it suddenly appeared in 2013. The basic technology has been around for years. It’s just that 2013 is the year that rapid expansion started to happen. We now have three standardized wireless platforms that are thoroughly disseminated—Wi-Fi, Bluetooth and 15.4 (Low-Rate Wireless Personal Area Networks). And we have ‘smart’ embeddable devices that we can now stamp out in the billions.

“Silicon is cheap, so it’s kind of a no brainer to install software into objects. In short, we can now move quickly to an intelligent environment.”

Like Franklin, Culler is most excited about the efficiencies the Internet of Things can bring to bear on an unruly and profligate world.

“I’m very hopeful that it can lead to huge cuts in waste in transportation, logistics and construction,” he says, “and it could really help spur sustainable development. For example, a building system could adjust its power consumption to exploit renewable energy. We now have the ability to differentiate ‘brown’ electrons—electricity from standard high-carbon sources—from ‘green’ electrons, or electricity from sustainable sources. Buildings could pick and choose among those electrons.”

The movie Wall-E foreshadowed a possible end point: All of us lying around obese and immobile on floating divans while flying robots endlessly succor us with liter-size containers of fizzy, flavored syrup.

UC Berkeley has been at the forefront of the Internet of Things, including major contributions to the development of Smartdust—infinitesimal micro-electromechanical sensors, some only a few square millimeters in surface area, which can be disseminated over wide areas.

“Smartdust was—and for some applications such as national defense, still is—a very interesting approach toward an intelligent environment,” Culler says.  “But as (the Internet of Things) worked out, it wasn’t the best approach. It ultimately was easier just to make familiar things smarter. It became apparent that making a thermostat so small you couldn’t see it didn’t make much sense.”

It may seem that the greatest danger of a smart environment is making the people who live in it passive and lazy. The movie Wall-E foreshadowed a possible end point: All of us lying around obese and immobile on floating divans while flying robots endlessly succor us with liter-size containers of fizzy, flavored syrup.

But Greg Niemeyer, a Berkeley associate professor of art practice, suggests that the Internet of Things could lead to productive citizen action—as long as we seize the opportunities and fight self-delusion. His own work provides some direction in this regard. In 2008, he kicked off the Black Cloud, a MacArthur Foundation- sponsored project aimed at improving air quality. That schema ultimately transmuted into a start-up company, Aclima Inc.

The Black Cloud/Aclima process involves measuring five air quality parameters through a vast number of locally-deployed sensors that constantly transit data on changing air conditions. Ultimately, Black Cloud arrays could allow for the precise real-time pinpointing of minute quantities of atmospheric pollutants.

“I’m a culturalist, not a technologist, so my main concern isn’t necessarily the data itself—it’s how we pay attention to the data,” says Niemeyer. “The Internet of Things is a perfect system in a world without humans. But people complicate things. Drawing qualitative judgments from quantitative information can be dangerous. Different people draw different conclusions based on cultural biases. So our greatest challenge is trying to figure out which questions we should ask, what queries we should make once we have the data.”

California magazine tackled the wrangling of Big Data in its Winter 2013 issue. This article by science editor Anne Pinckard takes readers inside Berkeley’s AMPLab, which was funded for six years beginning in 2011 and is devoted to making better use of the massive amounts of information generated and aggregated in the world today.

Share this article