Blazing-fast speeds! Zero latency! Moar data to moar devices! Unless you’ve been trapped in a tech-news dead zone, you’ve heard that the rollout of the next generation of wireless broadband has begun. Still, smartphone data addicts shouldn’t hold their breath for speeds of 10 gigabits per second. To provide the kind of 5G coverage consumers will expect, carriers will need to install as many as 20 access points per square kilometer, an expensive endeavor that will take years. Until then, we’ll have to accept that 5G is here, but it’s unevenly distributed. Here are some places to watch for it in the (nearish) future.
Cars Will Flow Like Schools of Fish
The folks who are gunning to make cars drive themselves are itching for 5G connectivity. Why? The faster you can get data into and out of a rolling robot, the better the experience. Constantly updated, ultrahigh-res maps of their environment make the ride safer and smoother. Developers in remote operation centers will also be monitoring lidar and camera feeds to keep an eye on their creations. And, of course, while they roll, their liberated occupants will demand streaming entertainment (and advertisers will demand to pummel them with targeted ads). But that’s all for the current kind of self-driving car, the one that watches but doesn’t talk to its surroundings. Way more exciting, if we’re talking real 5G, is not replacing human drivers but completely rethinking the way cars drive.
Link vehicles together and we’ll solidly surpass human limitations. Cars could move like schools of fish, in unison, smoothly and tightly, without colliding.
Engineers have longed to let cars swap data on location, speed, and heading for decades, and since the ’90s many have pinned their hopes on short-range radio transmitters. In 2017, UC Berkeley researchers sent a trio of connected semi-trucks down a highway with just 60 to 140 feet between them. Such convoys could improve fuel efficiency by letting vehicles draft each other and might even allow for going human-free in all but the lead truck. But the tech’s max range is only about 3,000 feet, it can’t handle many vehicles at once, and it requires special hardware in each car. Which helps explain why such luxuries are available only in a top-line Cadillac—leaving precious few chances for meaningful carversation.
But the anticipation of 5G has tickled engineers’ minds, enabling a new approach called “cellular vehicle to everything.” CV2X lets vehicles, infrastructure, and anyone with a cell phone link up over short distances and tap into cellular networks for long-range transmissions. Plus, many automakers are already putting wireless modems in their rides. (Then again, all these wireless connections will be irresistible to hackers, posing a formidable security challenge.)
In a demo last year, Audi, Ducati, and Ford used CV2X to warn drivers about oncoming vehicles that were outside their line of sight. A recent test by Ericsson, Qualcomm, and other companies helped cars merge smoothly and safely onto a highway, not just by communicating with one another but by taking orders from a central control system that worked like an omnipotent traffic cop. That central authority, enabled by 5G, underlines engineers’ high hopes for CV2X: Cars that can talk to each other will stop hitting each other and will warn each other about hazards ahead. Cars that can listen to directions will start traveling in schools, maybe even turning roads that aren’t getting any bigger (and shouldn’t!) from clogged arterials to free-flowing speedways.
With data flying between cars, then, it isn’t hard to imagine the next curious change. We’d need fewer of the algorithms that manufacturers are so madly uploading into today’s self-driving vehicles. And like most kids who think they’re too cool for school, our current cars of tomorrow might quickly be left behind. —Alex Davies
Can You Heal Me Now? Surgery Goes Wireless
The catheter inched down the middle-aged man’s coronary artery. As he lay on the operating table at Apex Heart Institute, a hospital in Gujarat, India, the instrument inflated a tiny balloon to widen the blocked vessel and installed a stent to keep it that way. The angioplasty went off without complications, but there was one big anomaly: The surgeon was not in the room, or even the building.
In fact, Tejas Patel, Apex’s chief interventional cardiologist, was some 20 miles from the patient. In the world’s first remote human heart surgery, he manipulated a teleoperated robot via joysticks at a makeshift workstation in a Hindu temple, chosen for its unmistakable spiritual significance and reliable internet connection.
The angioplasty wasn’t the first-ever surgery via the internet: In 2001 two surgeons in New York extracted a person’s gallbladder in France. Nonetheless, most robotic procedures these days involve patient, surgeon, and robot in fairly close proximity. But as connectivity improves and mobile networks get faster and less laggy, some startups and surgeons think it’s time to make internet surgery a routine option. “This technology is going to eliminate distance between doctors and patients in underprivileged areas,” Patel says. “I will love to do this procedure transcontinentally.”
The angioplasty in Gujarat was one of five, all successful, performed using a robot called CorPath, built by Boston-area startup Corindus Vascular Robotics. (The company was recently acquired by Siemens Healthineers for $1.1 billion.) Corindus started selling its bot in 2012, for use by a surgeon at a patient’s side. But a few years ago its leaders began investing heavily in telesurgery. The bet was that internet and wireless networks could let surgeons phone it in from almost anywhere—if carriers could fix the occasional glitches that mar video calls. “You don’t want that if you’re inside someone’s heart or brain,” says Corindus CEO Mark Toland. “You have to feel that you’re in the room.”
Internet connections are now looking reliable enough. In Gujarat, the procedure was performed over a fiber-optic link; in the US, Corindus is testing surgeries over 5G mobile networks. In one recent trial, a doctor in Boston steered a catheter through a surgical simulator, complete with digital beating heart, laid out on a table thousands of miles away at a test facility in San Francisco. Imagine: If doctors were able to perform remote emergency procedures to treat heart attacks or strokes on, say, military vessels, or in parts of the world like rural India that have few heart specialists but are relatively rich in mobile coverage, a lot of lives could be saved. “There are millions of patients who die—or live with severe problems—because treatment was not available on time,” Patel says. Surgeons won’t all become remote workers overnight, and internet infrastructure evolves slowly, but with 5G, eventually, there might always be a doctor in the house. —Tom Simonite
Assembly Lines Powered By Data—and Fewer Humans
At Foxconn’s megafactory in Shenzhen, China, thousands of young people shuffle between cramped dorms and monotonous production lines. But in one vast space, where green lights glow atop rows of humming equipment, robots ferry parts between machines, mechanical arms grab and place widgets at superhuman speed and precision, and cameras inspect circuit boards for defects. Few humans are present.
This is a new kind of assembly line, and it may someday put many humans out of a job. Behind all the automation is a tsunami of data. The machines send incredible amounts of information at astonishing speed—every detail of their behavior and performance—to 5G transmitters dotted around the building. It’s called Industry 4.0, and it promises to spark a revolution in productivity.
Citing rising wages and a tight labor market, Foxconn has added ever more automation to its factories. And faster, more powerful wireless tech will help choreograph the increasingly complex dance between robots and human workers. Of course, assembling an iPhone with robots remains a challenge—human fingers are still superior at manipulating fiddly electronics—but 5G will inch production closer to that goal.
Foxconn Industrial Internet, a spinout of the manufacturing giant, is using 5G to provide a real-time picture of the assembly line. The sensor data from each machine could let, say, Apple engineers in California monitor production of the latest iPhone, allowing them to make tweaks to boost output or fix a defect in minutes instead of days.
When assembly bots are able to share their data, a factory’s customers will also be able to take better advantage of machine learning. Fed gigabits of data per second, an AI thousands of miles away could unearth signals that point to potential problems (imagine a motor that’s wearing out) well before they occur—like a minority report for robots.
Foxconn isn’t the only manufacturer keen to take advantage of the new wireless technology. With sensors, 5G, and machine learning, “we can predict a robot failure three months in advance,” explains Michael Raiford, a VP at Samsung Semiconductor, which recently partnered with AT&T to set up a 5G network at a chipmaking facility in Austin, Texas. They’re testing wireless sensors that will monitor workers’ vital signs for indications of accident or illness—humans wearing out.
Not everyone sees 5G as a revolution. “People think it’s this big turning point,” says Willy Shih, a Harvard professor who studies manufacturing, “but I think it’s kind of a natural evolution.” In other words, 5G is simply another adaptation to a world that ceaselessly demands more data, faster. —Will Knight
The Next Virtual Reality Will Become the New Reality
Even if you’ve got every virtual reality headset ever made—HoloLens, Magic Leap, exfiltrated prototype of Apple’s long-rumored smart glasses—you haven’t experienced the best that AR and VR have to offer. Not even close. That’s because the magic is stuck in labs, where the computing is exponentially more powerful and the gigabit wireless networks plentiful. Headsets that look back at you and reproduce your face in bits; real-world environments digitized in real time so users miles apart can share the same space: With 5G, these projects will finally burst out of the pipeline and into your eyeballs.
Granted, the hardware needs to get better, from optics and battery life to thermal management, but there’s only so much VR any device can handle at 4G. Now, once you start pulling down data at speeds more than 100 times faster than your current phone, wireless headsets will be able to render VR representations that look exactly like you, doppelgängers that vault across the uncanny valley and ape your features and tics as you talk. It’s how you’ll be able to peer through your glasses to see not the windowless room you and your colleagues are sitting in but the sun-drenched quarters of a Swiss ski chalet. Today, you call your mom on FaceTime. Tomorrow she’ll see hologram-you in her real kitchen while you see the virtual version of all of it: Mom, kitchen, and the new router she wants your help installing. No, the cable goes there!
It doesn’t stop at familial IT duty. The mirrorworld, that global layer of data driving augmented reality, will finally leap past Pokémon Go and Minecraft Earth to become a ubiquitous, useful infrastructure: AI-powered avatars (some corporate mascots, others puckish indie counterbalances) will populate public spaces like discreet concierges, supplying everything from directions to sightseeing tips. You’ll capture photos and videos that aren’t just flattened images but places, then share them with friends so they too can stroll through the secluded Roman piazza you just explored. This isn’t an information superhighway, it’s a Super Mario Bros. warp pipe, big enough and fast enough to make the entire world broadcastable at the exact speed of experiencing it. It won’t be all fun and games, though. There’ll need to be robust antiharassment and safety tools built in—because if today is a picnic for trolls, it’s up to us to make sure a virtualized future isn’t their goddamn paradise. —Peter Rubin
