Ah, and here it is.
Far down the hospital hall, double doors part to reveal the
The robot, known as a Tug, edges closer and closer to me at the elbow of the L-shaped corridor and stops. It turns its wheels before accelerating through the turn, then suddenly halts once again. Josh, the photographer I’d brought along, is blocking its path, and by way of its sensors, the robot knows it. Tug, it seems, is programmed to avoid breaking knees.
This hospital—the University of California, San Francisco’s Mission Bay wing—had opened four days before our visit. From the
The whole circus is, in a word, bewildering. The staff still seems unsure what to make of Tug. Reactions I witness range from daaawing over its cuteness (the gentle bleeping, the slow-going, the politeness of stopping before pancaking people) to an unconvincingly restrained horror that the machines had suddenly become sentient. I grew up in Silicon Valley and write for WIRED and even I’m confused about it. The whole thing is just weird.
It’s really weird. And I’m not sure I like it much.
Roll, Roll, Roll Your Scary-Intelligent Medical Robot
The Tug that’d emerged without so much as smoke or pyrotechnics had come from the kitchen, where the“We’ve named ours after fruit,” he says, forcefully, over the fans. “So we have Apple, Grape, Banana, Orange, Pear—and Banana is out right now. At some point we’ll get them skins so they actually look like the fruit.” Other departments have their own naming conventions, with monikers that include Tuggy McFresh and Little McTuggy, plus Wall-E and of course the love of his life, Eve (the hospital is apparently trying to get permission from Disney to dress them up like they appear in the movie). Other Tugs will be stylized as cable cars, because, well, it’s San Francisco and why the hell not.
If you’re a patient here, you can call down to Henroid and his team and place your order if you’re keen on being a savage, or you can use the fancy tablet at your bedside and tap your order in. Down in the kitchen, the cooks—who aren’t robots—fire up your food, load it onto a Tug, and use a touchscreen next to the docking stations to tell the robot where to go. Once the food is loaded, the Tug will wait for 10 minutes, then depart, whether it has just one tray or 12, its max capacity.
There are no beacons to guide the Tugs. Instead, they use maps in their brains to navigate. They’re communicating with the overall system through the hospital’s Wi-Fi, which also allows them to pick up fire alarms and get out of the way so carbon-based lifeforms can escape. Rolling down the halls using a laser and 27 infrared and ultrasonic sensors to avoid collisions, a Tug will stop well away from the elevators and call one down through the Wi-Fi (to open doors, it uses radio waves). It’ll only board an elevator that’s empty, pulling in and doing a three-point turn to flip 180 degrees before disembarking. After it’s made its deliveries to any number of floors—the fleet has delivered every meal since the hospital opened—it gathers empty trays and returns them to the kitchen, where it starts the whole process anew.
And the cooks and other kitchen staff, says Henroid, adore them for it. “In fact, I think the most interesting thing is people have been very respectful of the robots. When we went and talked to other people at other hospitals, they said, ‘Oh, people get in the way.’ We haven’t had any of that. I think we did a lot as an organization to sort of prime people and say, ‘Hey, the robot’s got a job to do. Stay out of their way.’”
It sounds demeaning, but the humans had been coached on how to deal with robots. So welcome to the future. Your robot ethics instructor will see you now.
“Tuggy! Tuggy Tug!”
Isaac Asimov had three now-iconic rules for robots: They can’t hurt us or let us get hurt, they must follow orders, and they must protect their own existence. We can now tack onto these the new rules for the humans who interact with medical automatons.“We had to train on a lot of robot etiquette, you know,” says operations director Brian Herriot as we walk the halls in search of Tugs, aided by a laptop that tracks their movements. “Which is, we train them to treat a robot like your grandma, and she’s in the hospital in a wheel chair. If something’s in their way, just move it aside, don’t go stand in front of them.”
Asimov’s laws are good to keep in mind so we don’t end up with murderous hordes of machines, but we need to start talking more about the other side of things. How should we treat them? We need laws for human-robot interaction. For the moment, it seems that we’re supposed to just pretend they’re Grandma. That’s Law Number One. What the other laws will be, I’m not so sure. How will we treat AI that’s smart enough to pass as human, for instance? I mean, we’re already getting emotional about a box that rolls around hospitals. Maybe it’s too early to tell these things. Give me some time to think about it.
In this hospital, Law Number One is working. Most staffers have a strange nonreciprocal affection for Tugs. Reactions to our convoy of PR reps and technicians and me and Josh and of course robots included, but were not limited to:
• “Wall-E has an escort?”
• A woman watching a Tug turn: “I usually call it the Tug shuffle.” And her companion, subtly one-upping her with nice alliteration: “The Tug tango?”
• “Tuggy! Tuggy Tug!” And from a fan of brevity: “Tuggy!”
• Plus an outlier from two women who turned a corner to find themselves face to face with a Tug: “Whoa! The robot scares us!” The other woman didn’t say anything, but she didn’t defend the Tug either.
The affection is no accident. Aethon, Tug’s manufacturer, designed it to be comforting, and not in the sense that they avoided things like painting flames on it. It’s more subtle than that. The tone of that constant beep, beep, beep, for instance, was designed to alert humans without being so annoying that you want to ring Tug’s neck.
And then there’s the voice. Tug is chatty. Lest you worry that it’s broken down while waiting for the elevator, it assures you: “Waiting for a clear elevator.” Once it gets one: “Waiting for doors to open.” Tug warns you when it’s about to back up, and thanks you after you’ve unloaded its delivery. Its voice comes in either soothing male, soothing female, or super-enthusiastic Australian bro (have a listen below). Aethon had contracted with a client in Australia and decided to offer the voice track to other hospitals. Australians are famed for their friendliness, after all.
The robot, I’m told, is on its way. Any minute now you’ll see it. We can track them, you know. There’s quite a few of them, so it’s only a matter of time. Any minute now.
It may have an adult voice, but Tug has a childlike air, even though in this hospital you’re supposed to treat it like a wheelchair-bound old lady. It’s just so innocent, so earnest, and at times, a bit helpless. If there’s enough stuff blocking its way in a corridor, for instance, it can’t reroute around the obstruction.
This happened to the Tug we were trailing in pediatrics. “Oh, something’s in its way!” a woman in scrubs says with an expression like she herself had ruined the robot’s day. She tries moving the wheeled contraption but it won’t budge. “Uh, oh!” She shoves on it some more and finally gets it to move. “Go, Tug, go!” she exclaims as the robot, true to its programming, continues down the hall.
For as cute as Tug can be—and it pains me to say this—it’s also a bit creepy. There’s something unsettling about a robot that’s responsible for human lives tooling around with minimal commands. Maybe it’s that I occasionally felt like we were hunting wild animals, wandering around in search of Tug after Tug. While technicians can track a Tug’s movements, it isn’t always easy to immediately pinpoint and intercept them. We’re both roving parties, after all. We’d turn a corner and expect to see a Tug, only for it to pop through a door seconds later. That accuracy ain’t too bad in the grand scheme of things, but it nevertheless instilled a kind of suspense. It was like tracking a deer that suddenly emerged from the grass … and started beeping.
Alright, fine, maybe that simile isn’t airtight.
I, Robot Drug Dealer
There are two models of Tug roaming the corridors at UCSF Medical Center. The one that hauls food and laundry and such is like a pickup truck. It has a thinner front and a bed in back, which people roll big cabinets onto. The second is more like a van, boxier with built-in cabinets. This is the drug-pusher.We’re in the hospital’s pharmacy now, meeting Wall-E and Eve. You can tell the difference because behind each hangs a plush toy of their namesake. They’ll hang there until the robots get their new outfits (pending approval from Disney’s lawyers, of course). A pharmacist gathers some drugs, scans their codes into a touchscreen next to the robots, and chooses the destination for each. Walking over to Wall-E, she enters a code on a number pad, then places her thumb on a biometric reader to unlock the machine. A small screen on the robot tells her which medication goes in which numbered drawer, and she proceeds to pop each open and place the drug inside. With a tap of the green button atop Wall-E, the robot is off.
I know at least a dozen of you are thinking that maybe you should get into the Tug drug heist business, so I’m gonna save you some time and embarrassment. Not only does unlocking the drawers require the PIN and thumb print of the doctor or nurse who requested the drugs, but Tug won’t unlock until it reaches its destination. Anywhere else and it’s sealed tight.
So Drug Tugs securely deliver medications, and Linen Tugs haul was much as 1,000 pounds of laundry, and Food Tugs deliver 1,000 meals a day. We might begin to wonder about the people who previously did all that scurrying about. What was their fate?
Well, according to Pamela Hudson, the medical center’s associate director of administration, their jobs are safe. In fact, she says that with such a massive new hospital, hiring in some departments is on the rise. The robots are about supplementing current jobs, she says, not eliminating them. “It would be a travesty for us to hire more techs who specialize in instrumentation but all they’re doing is running around delivering trays,” Hudson says. “That’s not the best use of their skills—that’s not a real job satisfier.” As an added perk, she says, if staffers aren’t pushing around huge carts, they’re not straining themselves or mowing down their colleagues.
Just down the road in Silicon Valley, El Camino Hospital has been using the bots since 2009. And according to its chief information officer, Greg Walton, there’s huge pressure to bring down the absurd cost of medical care in America, and Tugs have allowed them to avoid hiring additional staff. “So by being more efficient we’re able to devote more of our dollars toward paid employees at the bedside caring for patients,” he says, “as opposed to pushing trash carts or linen carts or moving products and supplies throughout the facility.”
It’d be laughably optimistic, though, to say that robots like Tug won’t infringe on more and more jobs as they grow more and more sophisticated. It’s already happening elsewhere. Robots, long just stealers of manufacturing jobs, are breaking out of the factory into the world. There’s a hotel opening this summer in Japan with robot receptionists. Last week a Roomba ate a woman’s hair as she slept on the floor, which never would have happened had she hired a maid. Soon enough our taxis will drive themselves. And before long Tug will get smart enough to really start chipping away at the hospital workforce. When that happens, there won’t be an outfit cute enough to keep it from playing the villain.
In the Future, Robots Will Be Even Smarter and I’ll Still Be a Dum-Dum
Listening to my audio recording of the visit, there’s a period of about 10 minutes when every so often someone giggles. I hadn’t noticed it at the time, but there’s definitely some suspiciously frequent snickering there. And people seemed to pause before answering my questions, as if over-contemplating things. But these were patently easy questions.Riding an elevator to intercept another Tug, Josh points the camera in my face—and then it hits me. I hadn’t removed my fluffy white hairnet. Walking around a hospital in scrubs is perfectly normal, but wearing a hairnet beyond the kitchen is considered antisocial at best. I rip the thing off my head, and there is much laughter.
“I forgot about my hat. Thanks for telling me, guys.”
“Well, he told ya … with the camera,” someone in the convoy replies.
I’ve spent the morning tailing an autonomous robot that performs its duties without a hitch almost 100 percent of the time. And here I am, totally incapable of not making an ass out of myself in the line of duty. Right now I’m envying Tug not only on account of its perfection, but because it’s not programmed to feel embarrassment. All it does is roll around as doors magically part for it and doctors and nurses scurry about so as to not hinder Its Holiness the Tug.
Maybe that’s why super-intelligent robots make us uncomfortable. It’s not just fear that they’re dangerous or are going to steal our jobs—it’s envy. They’re not saddled with emotions. They do everything perfectly. They roll about assuming they can woo us with cute beeps and smooth lines like “thank you.” I, for one, shan’t be falling for it. I don’t like Tuggy one bit.
I throw the hairnet in a waste bin and continue on in search of the next ever-elusive Tug. It’s out there somewhere, helping save lives or whatever, trying a bit too hard to be liked. Someone’s probably calling it Tuggy Tug at this very moment, while I’m here trying to salvage what little social currency I have left.
There’s no robot for a man like me. Well, until I end up as a patient here, where there’s plenty of robots for a man like me. Then I’ll have no choice but to sit back and soak in the automated future of medicine—the beeping, the incessant politeness, the whir of electric motors. Count me out, though, when one of them starts talking like an Australian.
You can only push a man so far.
I petitioned unsuccessfully to get Josh fired for taking this photo. So he went ahead and added it to the story. Josh Valcarcel/WIRED