By 2020, there will be tens of billions of data-spouting devices connected to the Internet. And they're already changing how we live and work
IBM’s Almaden lab is sacred ground for techies. Set in the middle of 700 grassy acres on a hill south of San Jose, its scientists have filed thousands of patents. They’ve won Turing and Nobel prizes. And almost 60 years ago, they pioneered the first bulky disk drive. Since then, they’ve been involved in successive pushes to miniaturize it and miniaturize it again, so that now even the tiniest of devices can gather and store data.
Today, Big Blue is putting that tiny technology to work, developing a multi-application gas sensor that could help airports detect and track biochemical threats, determine whether the steak in your fridge has spoiled, or even diagnose breast cancer and other diseases simply by analyzing your breath.
Sensors like these are driving a relatively new tech trend: the Internet of Things. In essence, the “things” referenced in this awkward buzz-phrase are machines embedded with sensors that gather, store and analyze data. And since they’re all linked to the Internet, they can upload that data for further processing, download updated software and often be controlled from afar.
The international research firm Gartner estimates that by the end of last year, there were 3.8 billion connected things out there—smart cars, smoke detectors, door locks, industrial robots, streetlights, heart monitors, trains, wind turbines, even tennis racquets and toasters. By 2020, Gartner estimates there will be 25 billion of these smart devices, transmitting tiny amounts of data to us, to the cloud and to each other. Cisco’s outgoing CEO, John Chambers, has boldly proclaimed that there will be 50 billion devices online within five years, with a total market worth $19 trillion (all currency in U.S. dollars). Another leader in this sphere, Siemens, has said these smart things are starting to power a fourth Industrial Revolution (after steam, electricity and wired computers).
Some of this, of course, is hot air. There’s always a certain amount of hype that accompanies the latest tech trend. Remember radio-frequency ID tags? They were going to change the world, too.
And so, yes, there are technological obstacles between us and this Brave New World.
First off, there’s the language barrier. Smart home devices—one of the more developed realms within the Internet of Things—currently speak a Babel of wireless languages, depending on the manufacturer. Your home’s thermostat and HVAC system might communicate in Bluetooth, the fridge and coffeemaker in ZigBee, the locks and blinds in Z-Wave and the smoke detector in WiFi. Plus, making sense of the data produced by these machines—not to mention finding space to store giga-, tera-, exa- and even zettabytes of it—poses a huge challenge. Security is anoth-er ongoing concern. One IT expert recently demonstrated how easily he could hack into a radio-frequency-controlled insulin pump and remotely administer lethal doses to a diabetic. Other experts have claimed that hackers might, if motivated, access the software in smart cars to take control of their speed, brakes and steering.
But the tech challenges, as daunting as they are, worry insiders less than the legal, social and regulatory ones. Because these sensor-embedded machines will dramatically increase what we can find out about one another, they could give rise to so-called Big Brother and Little Brother problems.
First off, who owns all this data, anyway? Does the deeply personal information collected by your fitness tracker belong to you or to the manufacturer? Should law enforcement be able to access vehicle data in a criminal investigation? Will car-owners want their cars to alert authorities and insurers automatically after every fender-bender? New Cisco-made sensors can, when worn by miners, detect the presence of life-threatening gases. Others can sense if workers (say, in the oil fields or on mega-construction sites) are moving or still—perhaps injured. But they can also help employers determine precisely how and where their employees spend every working moment.
Even with such thorny problems looming, these intelligent machines are already altering spheres as diverse as health care and manufacturing, city planning, transportation and power generation, agriculture and household management. The devices themselves might be micro, but they’re causing macro shifts in how we live and work.
For a decade or more, consumers have been promised a future filled with fridges that text us when our milk goes off and coffeemakers we can turn on from our beds. But, though smart appliances have dominated the past several years of the Las Vegas geekapalooza that is the Consumer Electronics Show (CES), no singular product or dominant brand has yet emerged. Insiders describe this fast-growing market, which the firm RNR Market Research estimates was worth $20 billion last year, as a vibrant chaos.
The breakout success to date, particularly among the many earnest products aiming to reduce consumer electricity bills, is the Nest Learning Thermostat. With a clean design, a simpler interface than existing programmable thermostats and backing from Google (which paid $3.2 billion for the start-up in 2013), Nest has been making dramatic inroads in Canada, the U.K. and the U.S. Thermostats control more energy in your home than is consumed by your appliances, lights, TVs, computers and stereos combined. Nest boasts that its device—which “learns” your schedule, programs itself and can be controlled from your phone—could save users 20% on energy.
Other products are more in the novelty line. One of the hits of this year’s CES was Tagg, a way of remotely tracking the location and vitals of your dog or cat. Amazon announced in March a line of branded buttons you can stick around your house to enable you to order staples like laundry detergent and toilet paper with one push. Then there’s Brad, a smart but needy toaster that checks in with other toasters in his network to see how much action they’re getting and wiggles his toggle if he feels neglected.
There are some legacy companies providing whole-home solutions. New Jersey-based Crestron sets up bespoke systems so clients can manage their security, energy use, lighting, HVAC and entertainment systems from their tablets. And Ingersoll Rand has lately begun offering a cheap suite of home-control products under the Nexia brand.
Then there’s Apple—because, of course. It has announced the release of HomeKit, a platform it hopes developers in the smart-home sphere will use to create device-controlling apps. The aim is to provide one gateway, one lingua franca, for the industry. Apple generally only ventures into markets with huge potential, and indeed, RNR figures the smart-home market could grow to $60 billion by 2020.
Changes are coming to the office building, as well. No frippery here—these measures are geared toward bottom-line efficiency and worker convenience. Cisco, for instance, controls the core functions of its 300 buildings worldwide, including climate, electricity use and security, from four locations. The company foresees a day when an executive driving into the garage will automatically signal the elevator to come pick her up and turn on the lights in her office.
Buildings with sophisticated internal climate controls have become markedly more common. Manitoba Hydro’s new skyscraper in Winnipeg, for instance, features a massive natural humidifier—a steamy room, several storeys tall, filled with tropical plants and a water feature—connected to pipes that pump moist air throughout the building. The system knows when to open and close the blinds, either to let the sunshine in (to benefit from free solar heat) or to keep it out.
More than half of the world’s people now live in urban centres, and almost two-thirds of us will do so by 2050—which means 2.5 billion more city-dwellers to house, employ and transport. That’s a nightmare scenario for today’s cities, plagued, as so many are, by traffic, smog, crime, overflowing trash bins and inefficient lighting that gobbles between one-quarter and half of municipal electricity budgets. But technologies being tested right now will help the cities of the future better cope with the looming migration.
Stoplights with embedded video sensors can adjust their greens and reds according to where the cars are and the time of day. They’re a double-win, reducing both congestion and smog, since vehicles idling at red lights burn up to 17% of the fuel consumed in urban areas.
In Barcelona's Born Market, sensors embedded into parking spaces relay real-time information on empty spots to an app for would-be parkers. Siemens recently gave a grant to a start-up devoted to building parking drones that could guide cars to available spots. Sound trifling? It’s not: Up to 30% of congestion is caused by drivers cruising the streets in search of a place to park.
Tel Aviv is tackling traffic on busier roadways by reserving one lane for buses, shuttles, taxis and car poolers—and allowing impatient and deep-pocketed commuters to use the designated lane, as well. Sensors in the asphalt pick up the car’s licence plate number and automatically charge the owner’s credit card at a rate that varies depending on how busy the road is.
Smart LED streetlights in San Diego turn on only when a pedestrian or vehicle approaches—the city recently replaced 3,000 old streetlamps with sensor-equipped ones to save an estimated $250,000 a year. The Brits, in an effort to deter hooliganism, are testing a lamp that comes on extra-bright when it detects banging and hollering, and is armed with cameras that transmit a live video feed to the cloud.
Over in Philadelphia, they’ve invested in $4,000-apiece solar-powered garbage cans (called Big Bellies) that crush waste and send a missive to a dispatcher requesting pickup when they’re full. Philly has been able to reduce the number of weekly garbage-collecting shifts from 17 to just three, and realize $1 million a year in savings on fuel, maintenance and labour costs.
The Harley-Davidson motorcycle plant in York, Pennsylvania, was built in 1973 as a typical assembly-line operation. But six years ago, it got a high-tech update courtesy of Cisco. Now, a slew of sensors linked to so-called manufacturing execution systems collate data from the factory floor to pinpoint any bottlenecks. When a rear fender was found to be holding up the process, managers shifted the layout so that the parts flowed directly onto the line, rather than being gathered and moved manually. In another room, sensors can tell whether the conditions—air flow, moisture—are optimal for painting and amend them as necessary. The system didn’t come cheap. One analyst recently told The Wall Street Journal that installing a manufacturing execution system into a single factory can cost between $500,000 and $1 million. But according to SAP (which provided the Harley plant’s software), the factory can now turn out 25% more bikes with 30% fewer workers. Instead of delivering one of its 1,700 bike variations in 21 days, it can manage delivery in a mere six hours.
Over in Germany, Siemens’s plant in Amberg churns out close to 12 million programmable logic controls a year. (A PLC is a switchboard that can control systems as diverse as cruise ships, ski lifts and, yes, assembly lines.) Microsensors embedded throughout the manufacturing and assembly process have helped the company virtually eliminate defects: It claims its PLCs are perfect 99.99885% of the time.
Greater “visibility”—lingo for instant information delivered from sensors to smartphones and tablets—is also resulting in far less downtime for machines, since managers can detect bottlenecks and existing or impending maintenance issues before they blow up. Accenture recently reported that Internet of Things technology can trim average repair costs by 12%, maintenance by 30% and downtime by as much as 70%. It can also save big on electricity costs—one of a factory’s leading expenses—through smart energy management systems. Power rates are automatically incorporated into machine work schedules, allowing the plant to avoid peak prices.
As GE chairman Jeff Immelt once put it: “If you went to bed last night as an industrial company, you’re going to wake up this morning as a software and analytics company.”
The new Tesla S electric sedan retails for $70,000. In addition to its enviable pickup—zero to 100 kilometres an hour in less than four seconds—it can travel up to 435 klicks on a single charge. If you’re getting low on juice, the car’s navigation system can lead you to the nearest charging station. Elon Musk’s latest-generation car comes equipped with Autopilot, which uses a combination of camera, radar and 360-degree sonar sensors to automatically drive on open roads and in stop-and-go traffic, and to not only find, but back into, parallel parking spots. The camera also reads posted speed limits and can warn drivers to slow down. Veer out of your lane and the driver’s seat shakes.
And the Tesla S gets better all the time, thanks to internal Internet-connected software that sends a steady stream of data back to the company’s engineers. Since the car’s release, programmers have upgraded the software several times—to boost its range, enable it to warn the driver when other vehicles are sitting in their blind spots, and to automatically dim the high beams when another car approaches.
All this is to say that the days of your car being a static thing, its functions set at the point of sale, are receding fast. Legacy companies like Mercedes-Benz are also set to launch smart vehicles. At its Silicon Valley research centre, a team of engineers and programmers are perfecting a model that can interact with a smartphone, gathering information on your appointments and proposing routes to get there, and displaying real-time traffic information. In May, Freightliner (a subsidiary of Daimler, along with Mercedes) received a Nevada licence for the world’s first robo-truck, which had already steered itself through 15,000 kilometres of testing on the state’s roads (albeit with a human operator along for the ride).
The research firm Gartner has estimated that, by 2020, there will be 250 million connected cars on the world’s roads, with many of them capable of driving themselves. There are eight million traffic accidents each year and 1.3 million crash-related deaths; Cisco’s Smart, Connected Vehicles division has posited that autonomous cars could eliminate as many as 85% of head-on collisions. They could also help ease traffic, since they’ll be able to communicate their positions to each other and therefore drive much closer together than vehicles piloted by humans. Traffic experts call this “platooning”—packing more cars into the same road space—and it could help save drivers at least some of the 90 billion hours they currently spend stuck in jams each year, generating 220 million metric tonnes of carbon-equivalent and wasting at least $1 trillion in fuel costs and lost productivity.
The aviation industry has always been slow to adopt new technologies—America’s air traffic control system, for instance, still runs on rickety computer infrastructure built in the 1970s. And it’s understandable, to some degree: The consequences of a technological glitch can be particularly dire when we’re talking about mass movers 30,000 feet up.
But the fact that the technology to prevent tragedies like the Germanwings crash—in which a disturbed co-pilot locked out the captain and deliberately flew into the French Alps—already exists makes it all the more senseless. Airplanes have long been equipped with sensors that collect data on fuel efficiency, altitude, location and maintenance issues. But that data has typically only been processed after the aircraft lands. With advances in connectivity and data processing software, there’s no reason it can’t be sent off and parsed mid-flight. That same technology could also be used to override the pilots in crisis situations like Germanwings, or to up-load each plane’s location more frequently—which would have helped greatly in the search for Malaysia Airlines Flight 370, another re-cent aviation tragedy.
But change is coming, slowly. Sensors in an aircraft’s engines can now detect and isolate developing problems—in part by measuring the temperature of a jet engine’s exhaust—and communicate those to both pilots and ground crews while the plane is still in the air.
On the efficiency side, GE has developed a tool that measures fuel use inflight and subtly moves the wing flaps (among other things) to reduce unnecessary drag. The technology helped Alitalia reduce fuel consumption by 1% in a year. With industry-wide spending on fuel hovering at around $30 billion annually, even such small savings can add up.
The rail business, too, is slowly chugging toward modernity. Britain’s Network Rail Telecom and Cisco are in the process of installing sensors in and beside the tracks to inform a centralized command centre if they need maintenance, or are threatened by nearby landslides or flooding. This will reduce the estimated 1.3 million hours currently spent on rail inspections.
New York commuters have the Internet of Things to thank for shorter commutes. The city’s Canarsie subway line recently installed Siemens-made tracks and trains that can pinpoint location with far more precision than the old automatic block signalling system (which uses trackside lights to tell trains to stop and go based on when they pass fixed points). Because smart tracks know exactly where the trains are, that means the gaps between trains don’t need to be as large, allowing significantly more trains to run on the busy route—up to 26 per hour, instead of only 15.
Shipping is getting in on this, too. Germany’s main port of Hamburg—enabler of the country’s post-Second World War economic miracle—has faced several linked and bedevilling problems in recent years. Many of the 550 trucks that arrive at the port each day were idling for hours in long lines waiting for their ships to come in, or parking in residential neighbourhoods near the port, since harbour-side spots were scarce. With 10,000 ships unloading there each year, it often happened that too many arrived at once, jamming the relatively small harbour. Expansion wasn’t on the table, given the historic, highly populated nature of its location. Now, thanks to a project with Cisco and SAP, the ships and many of the nine million containers moving through the port transmit (and constantly update) their precise arrival times, so trucks can arrange just-in-time pickup and drop-off of freight. Truckers can even book parking berths remotely, so they don’t drive around looking for spots or clog other parts of the city by lining up.
Technology designed to help boomers live at home longer is, well, booming. This is hardly surprising, with almost 15% of Canadians now 65 or older—a proportion expected to rise to almost 23% within two decades. A new generation of sensors can tell if the condition of patients living at home has worsened—and communicate that at once to their health care teams. Philips—best known for light bulbs and electric toothbrushes—has created a pillbox that pops open when it’s time to take your meds, and sends a message to, say, a family member or nurse confirming that you’ve taken them. The Dutch company recently spun off a new health care subsidiary, Philips Healthcare, that is a leader in the field—and is struggling to find an interface that works just as easily for smartphone-wielding youngsters as it does for octogenarians with degenerative diseases and dementia.
Their sensors can be specially refined, like the ones used by neonatal units to monitor premature babies. Since they can’t be placed directly on delicate skin, the sensors instead use high-definition cameras to monitor skin colour, breathing and temperature, and alert nurses of any changes. These devices will eventually help doctors and nurses care for and monitor more patients both at home and in hospital beds. Smart beds now being used at New York-Presbyterian Hospital can tell immediately if a patient has gotten up, and let the nursing station know.
Then there’s the booming market for fitness trackers like the FitBit, Apple Watch, Suunto and others, which has already surpassed $2 billion, with well over 84 million sold so far. These monitors measure heart rate, sleep patterns, diet, exercise and more, and beam that data to mobile apps. Soon, that information could be sent directly to your health care provider or insurer, which still rely on your word that, yup, you exercise four times a week and always take the stairs. U.S. insurer John Hancock (a subsidiary of Manulife) is offering clients up to 15% off premiums if they willingly hand over data that proves they lead a healthy lifestyle.
Next up: subcutaneous implants. Mississauga-based Medtronic already sells a glucose implant that helps diabetics keep tabs on their blood sugar.
The grid was designed to deliver power on an as-needed basis, to delicately balance supply and demand—a challenge, given that demand varies by time of day, by weather and by season. A heat wave, a blizzard—heck, even an Academy Awards broadcast—can all stress this aged infrastructure. To meet sudden spikes, backup power stations and diesel generators must stand at the ready, gobbling up scarce resources. It is far from efficient.
The basic theory behind the so-called Smart Grid is simple: Power is priced on the basis of demand, and this information is transmitted immediately to smart meters, thermostats and appliances so that they can draw the power they need at off-peak times, when it’s cheapest. This system uses market forces to balance the system loads and should, in theory, make power networks less susceptible to black- and brownouts.
Pilot programs, most notably in Italy and Texas, have demonstrated that the theory can work in the real world. The U.S. has set 2030 as an informal deadline to implement most of the components of the smart grid; Ontario’s Hydro One is one of many regional utilities worldwide currently working to smarten up its network. It’s shooting for 2025, though it has already re-placed many old meters with smart ones.
For now, they’re simply transmitting time-of-day usage directly to the utility. But the meters could, in the future, receive information on pricing and the total demands placed on the system, and govern themselves accordingly.
Power lines and pipelines are getting a high-tech upgrade, too. Data collected by sensors in the lines can be analyzed to detect and isolate maintenance problems. And predictive software already on the market can anticipate which trees are most likely to fall and take down lines. Cisco builds pipelines lined with sensitive fibres that can sense leaks and radio for help right away. For aging pipelines, GE has developed software that collates seismic data, topographical details, population density, and hospital and school locations to help make maintenance decisions on an ongoing basis or in emergencies.
The growth of renewable energy sources also hinges in large part on the smart grid. By next year, according to the International Energy Association, renewables will replace natural gas as the world’s second-largest source of power (coal is still on top). Here in Canada, wind and solar are by far the fastest-growing power-generating sectors (though they still account for just a few per cent of the total). While they may be easier on the environment, they put major pressure on the grid, since the energy generated by solar and wind farms varies by time of year and day, throwing out of whack its delicate balancing act. Solar panels that can communicate the amount of power they’re generating already exist. It remains to knit fields full of these panels into the grid, and to find a scalable battery to store overflow when we don’t need it.
Wind is suffering similar integration issues, though the latest generation of turbines themselves are already benefiting from Internet of Things technology. GE-built turbines on the leading edge of a wind farm can let those behind them know that a gust is coming, prompting them to immediately alter the angle of their blades to protect themselves from damage and lengthen their lives. A relatively new software program also processes the data collected by turbine sensors and proposes the optimal angles to generate more power, increasing wind-farm production by as much as 5%.
Despite the bucolic image we might have of the average family farm, farmers have always been early adopters of technology—after all, anything that can help boost the meagre living they can scrape out of the land is a good thing. Most farmers these days walk their fields with GPS-enabled smartphones in their hands, loaded with ag-related apps. And with farms getting dramatically larger—the average spread in the United States has doubled in the past quarter-century—farmers (or, as is becoming more common, the huge corporations that own these operations) have been quick to deploy data-gathering, Internet-linked devices to help keep track of them. New machines from John Deere can not only plow, sow and reap, they can also collect a Farmer’s Almanac worth of data, including air and soil temperatures, moisture, wind speed, humidity, solar radiation and rainfall. Smart watering systems sprinkle just enough H2O on the fields, in just the right places, and can detect leaks in water pipes—vital in dry and drought-affected regions like California. One company has developed a sensor that can detect high counts of a particular pest and then release the pheromones that disrupt their mating rituals—which can, in turn, reduce the need for pesticides. Even cows are now transmitting bits of data in real time: A Dutch company has created sensors that, when attached to individual animals, can tell farmers which ones are in heat, pregnant or ill.