ParisTech Review

A visitor from 20 years ago would see many changes in the way we work, play, shop, and communicate, thanks to the Internet. Yet other aspects of our world have scarcely changed at all.

Our appliances, for instance, are mostly the same. Our home is mostly the same, but maybe with a bigger TV screen and a less bulky computer. Our transportation is mostly the same as well, despite the fact that new cars now contain up to 100 million lines of computer code and at least 70 microprocessors. Overall, the Internet has remained a network between human beings.

In a few years, however, that’s expected to change dramatically. At present, around 5 billion devices are connected to the Internet. By 2020, it could be 10 times that many. The president and CEO of Ericcson has told shareholders he expects to see 50 billion devices connected to the Internet by 2020. “Today we already see laptops and advanced handsets connected, but in the future everything that will benefit from being connected will be connected,” said the CEO, Hans Vestberg.

Brave New Gadgets
What will the machines be able to do for us that they haven’t already?

Some applications sound like science fiction but are fast approaching reality—or, rather, creating, in the phrase of some boosters, an “augmented reality.” For example, the Wikitude browser makes it possible to look through a smart phone onto a hyperlinked streetscape, where floating links identify buildings, tag street names, provide restaurant reviews, and more, based on which “world” of data the user downloads to overlay. Geopositioning information is overlaid with a variety of information from any number of sources, including Google Earth, Wikipedia, Booking.com, and archINFORM, a database of architectural data—any site that has data that can be tagged to a place. Wikitude is definitely a work in progress—on an iPhone in Paris, anyway, it seems to work in fits and starts—but like the Internet in 1994, the promise is clear.

Another application, Wikitude Driver, does something similar, but for drivers. This application may have a shorter life span, however, as it seems only a matter of time before the robots take the wheel. A team of Google engineers has developed an autopilot system for cars, using the same data and positioning technology that underpins Wikitude. So far, Google’s self-driving cars have logged more than 140,000 miles, including such famously curvy drives as Lombard Street in San Francisco and the Pacific Coast Highway. At this point, the biggest challenges to driverless driving seem less technical than legal: most authorities still stress that a car must be driven by a human being.

Other elements of this new world are already being rolled out: smart power-grid systems that monitor our energy consumption; dynamic, networked traffic lights that give the majority of drivers the best of all possible commutes; inventory management systems that make sure hospital equipment is never misplaced. Such inventory systems use radio-frequency identification tags (RFID), the radio-signal equivalents of bar codes, which can be tracked and used to generate a sort of real-life version of Harry Potter’s marauder’s map.

As the tags get smarter, they are beginning to include tiny microprocessors that can give not only an inventory number but broadcast their condition. At their most basic, these sensors will be able to answer questions about themselves, says Benn Konsynski, a professor of information systems and operations management at Emory University’s Goizueta Business School: “Where am I? How am I doing? What shape am I in? Where do I want to go?” Soon, too, they will be able to do a whole lot more.

Tech watchers speculate that this kind of embedded intelligence will make it possible for medicine bottles to report that they’ve been left in a high-temperature place for too long, for toilets to test urine and call the doctor if there’s a problem, for refrigerators to monitor their own energy consumption and alert the manufacturer to perform maintenance.

The technology is being used to tackle broader problems as well. Earth Networks, for instance, a Germantown, Maryland, company that delivers global weather and lightning observation services, is working with the Scripps Institution of Oceanography at the University of California at San Diego to set up the first global network of atmospheric sensors to monitor greenhouse gases and climate data.

Of course, as with any major technological shift, the Internet of Things will likely bring many other changes, good and bad.

Beyond general efficiency gains that should drive lower prices, consumers should expect to see their relationship with product providers changing. Where now the connection between manufacturer and user generally ends at the checkout counter, tomorrow the technology may lead to many new kinds of offers, such as paying for an appliance not at the check-out counter but per use, and ongoing monitoring and upgrades. “The critical thing is it changes commerce practice: contractual ownership, governance, safety, privacy,” Konsynski says.

Overall, the new technology should lead to closer relationships between buyer and seller, and even blur some of those boundaries, as more monitoring and feature upgrades become technically feasible.

“Is There Something Wrong, Dave?”
But those new relationships may come at a cost, warns Andrea Matwyshyn, a professor of legal studies and business ethics at the Wharton School at the University of Pennsylvania.

In particular, Matwyshyn and some other privacy law scholars fear that the new capacity for tracking things will spawn more corporate and government snooping, sharply reducing the size of the private sphere. The potential for monitoring and sharing highly granular information about all of us is growing, says Matwyshyn, and the power of the individual to control that information is decreasing. Society needs to decide “how that information can or cannot be collected, shared, and exploited,” she says.

It might not be like living with HAL, the malevolent, all-seeing computer in “2001: a Space Odyssey”, but some legal scholars worry that the growth of networked information could shrink our level of privacy. “Those zones of privacy that we’ve taken for granted can quickly be eroded without careful planning of the big picture,” Matwyshyn says.

A door you left open isn’t talking to anyone right now, she says, but hypothetically, a wired door could “inform your electricity or your gas company that you are a negligent door user.” Maybe the energy company scolds you about it—or maybe the information gets shared with the government. “You get labeled a careless, non-environmentally conscious person because the reason for your higher level of consumption of energy is arising from your negligence … and not some defect in your insulation that’s beyond your control,” she says.

It’s often noted that technical systems take on some of the ideological attitudes of the inventors. If that’s true, the truly global nature of the development of this technology may make the preservation of Western privacy ideas more difficult: The Chinese government has made the development of the Internet of Things an important national priority. Chinese Premier Wen Jiabao mentions it in many speeches and a 170,000-square-meter research center recently opened in Shanghai.

The To-Do List
Even if East and West don’t clash on privacy issues—and certainly, when Chinese officials talk about the Internet of Things, they tend to emphasize logistical applications, such as smart buses and smart insurance cards—other challenges remain.

Speaking for himself, Stephan Haller, an Internet of Things researcher at SAP Research in Zurich and a member of the European Union-China Internet of Things Expert Group, sees a number of issues still to be resolved.

First, how do you connect to a wide variety of devices, although they may not speak the same protocol, or language? In the development of the Internet so far, most of the connecting devices are PCs, and all have used the same Internet protocol.

Second, how do you deal with failed sensors? When tracking multiple sensors, how do programs handle sensors that send back unreliable data? How do you detect that the data is unreliable in the first place?

Third, who is liable if something goes wrong? The recent failure of the iPhone alarm clock program on Jan. 1 may not have been so serious, but problems with future networked systems might be. “If you have a machine doing something suddenly that it’s not supposed to be doing, who is liable for that?” Haller asks.

Matwyshyn also worries about “the unforeseen interactions of the various pieces of this new system that we’re building.” Often, it’s only after the system is built that problems emerge, she notes. One case in point: automated black-box securities trading, which now represents 70 percent of market volume, has been blamed for a number of trading irregularities in the last two years.

“Sometimes there’s an over-exuberance with respect to the capabilities of the new technologies that we’re building,” Matwyshyn says. “Fantastic as they are, they are still built by humans and for humans.”

Previous Post Next Post