I’m standing in a room in El Dorado, Kansas—pronounced el-doe-RAY-doe, population 12,852 as of 2013—looking at stone arrowheads. This is a small collection mounted in a handheld glass case, arranged in starburst pattern and presented for my inspection by Ardath Lawson, educator and librarian at the Kansas Oil Museum. Lawson, chipper and knowledgeable, sets the case down and gestures for me to follow her down the hall to the kafir corn exhibit, mentioning that the corn is a kind of sorghum crop popularly grown in the area before oil was discovered. On the way I stumble over a seam in the carpet and accidentally knock my face into the wall.
It’s a little embarrassing, but it doesn’t hurt, because it’s not my real face. My actual physical face is 1,357 miles away in New York, touring the Kansas Oil Museum though a Double Robotics telepresence robot. The robot I’m inhabiting is nicknamed Albert, after Albert Hogoboom, the departed patriarch of El Dorado’s Hogoboom Oil Field Trucking. Contributions made in Albert’s memory funded the purchase of the robot, and in his honor, the robot is painted bright red, just like the Hogoboom trucks.
I drive Albert through the museum as Lawson shows me photos of the Queen of the Kafir Corn Festival dating back a hundred years, before they started drilling for oil in 1915. I peer into a replica of a barn, and I spin Albert’s wheels on a tile-floored re-creation of the Butler County courthouse. I telescope Albert’s “neck” upward to get a better look at an austere mannequin dressed like an oil surveyor; he resembles a respectable gentleman on safari rather than a roughneck wildcatter.
The Kansas Oil Museum is a compact set of exhibits. But if I wanted, I could take Albert rolling over all 10 acres of the grounds for an extended tour. The museum installed WiFi coverage to make this possible so Albert wouldn’t be confined indoors. They’ve gone to a lot of effort, though so far only about a half-dozen tours requested Albert in his first year of service. I ask Lawson what made them think of offering robotic tours of the museum, and she tells me they got the idea from the Kansas Historical Society over in Topeka.
Sure enough, the KHS has its own robot, IKE (the Interactive Kansas Educator). Program coordinator Joy Brennan says IKE, like Albert, has only drawn a handful of tours so far. Both robots are pitched as solutions for cost-conscious Kansan school districts, offering something akin to a field trip at a fraction of the price.
Brennan says one thing common to these tours is students’ intense curiosity about what IKE looks like. Tour guides have learned to show a photo of the robot to itself, so the children can see through its eye what “they” look like, quietly trundling around in Topeka.
I had that same feeling the first time I drove a telepresence robot. My debut was at Double Robotics’ offices in Burlingame, California, south of San Francisco. (I was of course physically still in New York, looking through the robot via my laptop.) In the Burlingame office, publicist Sara Broyles gave me a few tips on driving the robot around a large, featureless, almost empty room. Then she directed me to look at “myself” in a full-length mirror on the wall. This need to examine your robot body must be a common urge in first-timers.
IKE and Albert, and mid-level telepresence robots generally, resemble iPads on sticks because that’s what they are. The minimal physical presence of the robot itself means you, and others, focus on the human image in the tablet screen rather than the robot. It’s certainly still weird, but the utterly nonanthropomorphic robot frame neatly avoids the uncanny valley problem. Of course, some people still get a little creeped out.
But the truly striking and unexpected thing about driving the Double was how immediately I had sense of being it—of being where it was, as a person. Perhaps this is an effect of first-person video games or the general normalization of digital culture, but there’s a very clear and perhaps unconscious sense of navigating the space where the robot is.
At the Double office, Broyles guided me into a side room, and there, suddenly was the ocean—their building is built on a shoreline, meaning that the floor-to-ceiling windows looked directly onto placid waves. I marveled at the view in a distinct way I recognized as being like I was there, appreciating it, rather than seeing it as a picture, or a video, which is what I was really doing.
I followed Broyles back into the big empty room with the mirror. As we talked, I noticed an open doorway at the farthest end. I attempted to distract Broyles by asking about something near the opposite end of the room, then activated the robot’s “turbo” feature—roughly human walking speed—and made a break for the open door. Broyles intercepted me easily and congenially enough that I suspect I wasn’t the first one to try an escape.
Telepresence robots have been around for a few years now, but they haven’t captured consumer interest like drones, or GoPro cameras, or similar gadgetry. Most people who encounter a telepresence robot will do so at a tech company, where they’re often tolerated as novelty gizmos or the indulgence of a geographically distant executive. There are arguments for telepresence in remote education and elder caregiving, but these haven’t caught on significantly compared to the office use case.
But there’s also a sustained level of broad interest and investment, globally—a feeling that telepresence can offer something not quite possible otherwise, if we can just figure it out. Telepresence robots are not just remote-control vehicles intended to let you do stuff remotely. They’re meant to project your actual human presence into a different locality, where you relate to other humans as a present human. If Facebook Spaces represents a simplified future for cartoony virtual reality, this is the opposite: clunky old-fashioned remote reality, complete with laggy video and janky audio as your connection surges and wavers. That’s what makes these robots both satisfying to use and odd to observe in action—the human-ness or not-human-ness of a robot version of a person, physically there.
Hard to say, though, how you turn that appeal into a moneymaker. You can, at present, spend as much or as little as you want on a telepresence robot. Double Robotics has enjoyed what passes for success in a relatively small market, selling over 8,000 robots since the company’s founding in 2013 for $20 million in revenue. The latest Double Robotics model, the Double 2, is still an iPad-on-a-stick. But it has a stout axle and solid wheels so it steers and balances better, along with attachments for better video and sound. It starts at $2,499.
Other iPad-on-stick robots range further and further down in price, all the way to the super-simple Clone at $499. (The iPad is not included in any of these.) Even cheaper, you start to get cute mini-bots like RambleBot for $199 (extra for claw attachment). On the other end of the spectrum, more formidable telepresence robots with custom screens and substantial torsos include the VGo (starting at $3,995), the Giraff ($11,385), and the BeamPro ($13,950). Roomba maker iRobot once had its own high-end telepresence robot line, but unhappy investors essentially forced its shutdown; two veterans of that effort received funding in 2016 for stealth startup that, sadly, seems more about traditional video conferencing.
BeamPro maker Suitable Technologies has led the charge since 2011 for telepresence as a calling card for the globally distributed business. Their legendary 2013 promotional video predicts a robot-infested future that is still strangely compelling despite not having come to pass, exactly.
Using a Double robot in my office, I noticed many of the same psychological dynamics as Seth Stevenson did when he reviewed an early Beam model for Slate back in 2014. At first it felt silly gliding around, and it seemed unlikely I could get any real work done without becoming distracted by the experience of robot-driving, or distracting my colleagues.
But then by pure coincidence, I was asked to hold an impromptu meeting explaining our choices for company health insurance plans, of all things. So I took up a position in a corner of a conference room, extended my robot’s neck to “standing” height (a commanding five feet or so), and discussed with a couple dozen mortal humans how they might preserve their precious, ephemeral existence a few moments longer through sensible health care.
And yes, it was weird, but astonishingly quickly, I forgot I was doing anything other than holding a meeting, answering questions, taking notes, and so on. When the meeting ended, a few more people sat down at the table for some follow-up, and I dutifully “sat” as well (lowering my “neck”) to join them. I rotated to “look” at people as they spoke in turn, and people looked at “me”.
Later, I asked people what it was like dealing with me in that form. Two said it was still weird and off-putting, but most responded with a shrug. It got normal, just in the course of one meeting.
Being in two places at once—or bilocation—is a psychological illusion of telepresence that people seem to accommodate with surprisingly little difficulty. That’s definitely the case where telepresence robots are in regular use, such as at the corporate headquarters of Suitable Technologies, makers of the Beam, where the robots are legion. “Sometimes, I’ll overhear people talking in the hallway outside my office,” says Bo Preising, chief strategy and product officer at Suitable, “and I’m not sure if they’re really there, or Beaming in.”
Preising, whose pre-Suitable background was in medical technology (including robotic surgery), recalls being annoyed by the limitations of video conferencing outside fixed locations of properly equipped rooms. “You’d have someone holding up a laptop in a factory or something, running Skype,” he says. “It was so painful, you’d finally just rather spend the money to fly out there instead.”
With a robot, the user can drop in and out without needing a local minder. You can make appointments or strike up impromptu conversations (about health insurance, for example). Preising says their research indicates people are 60-80% more engaged with a person communicating via telepresence than via traditional video conference or phone call.
Suitable is certainly bullish on the market potential of telepresence, and their competitors at Double Robotics are pleased enough with their prospects to seek Series A funding. Even while happy with growth in office usage, most vendors are looking for other markets too. A common scenario has relatives using a robot to “visit” someone in a managed healthcare situation like assisted living or hospice. Preising himself used a Beam to allow his family to visit his ailing mother, and he says that while some relatives found it difficult, others were able to get past the technology and have an emotionally positive experience—where repeated physical visits might have been prohibitively expensive or practically impossible.
David Cann, co-founder and CEO of Double Robotics, is less enthusiastic on this front. “We looked into using a Double as a home health aid, or for visiting elderly relatives,” he says. “But we found that people really wanted to actually be with their loved one, not just talk to them.”
It’s an unusually candid admission of the technology’s limitations, but Double Robotics tends to be more conservative about exploring new market niches. “Our use cases come from our customers,” Cann says. “We prefer more of a ‘pull’ approach, rather than ‘push’.” While several telepresence companies tout their robots’ educational uses for remote learning—parodied in popular culture in Bob’s Burgers and Community among other series—Double has pioneered one unique pitch in modifying their robot to serve as a camera platform for 360-degree video. This solves the common problem of having to hide the human operator of a moving 360-degree field of view, and came about from actual customers who had jury-rigged a Double for the purpose.
Telepresence robots have started to show up at technology conferences, which seems like a decent trade of personal immediacy to avoid the logistical nightmare of actually attending some of the larger shows. Surprisingly—to me anyway—is how slowly the industry is awakening to the potential of releasing telepresence robots from biz-corp confines. Double Robotics is not alone in having their units do museum tours; there’s a Beam (or “Moe-Bot”) at the Las Vegas Mob Museum, and similar Beam robot presences at other museums nationwide. Robot salespeople have started appearing in retail settings too; there’s a Beam store in Palo Alto staffed entirely by robots, and supposedly none have been abused or kidnapped to date.
Despite the varying robot physiques, a software play is integral to any telepresence business. Duy Huynh, CEO of Autonomous, views his relatively inexpensive Clone robot primarily as a platform. “Telepresence is just an app,” he says, seeing as much market potential for a fleet of otherwise algorithmically driven Clones serving as “an extension of awareness for a human operator.” Think of freely wandering robots evaluating their environment, like a server farm or warehouse, automatically; when they encounter an anomaly the program can’t parse, a human is summoned to take over a robot and investigate.
After working in my own office remotely for the usual meetings and conversations, I couldn’t resist taking my robot out for a spin in New York’s Soho and Chinatown (escorted of course). It’s hard to explain, but the experience was absurdly fun and weirdly stimulating. I drove down the sidewalks, conversed with a few passerby, zoomed down a vacant soccer field, shopped for Asian produce, and ordered a beer in bar where I could only gaze longingly at the frosty beverage forever out of reach.
I enjoyed the game-like aspect of piloting the robot in a real space, but I also really wished I could be where the robot was (and not just for the beer). That is to say, touring New York via robot is absolutely no substitute for the real thing. But if I could use a telepresence robot to “visit” places I can’t get to for reasons of time or expense, I would absolutely consider it. It would, if anything, enhance my desire to ultimately visit the real place having gotten so close, and yet so far.
Using telepresence for tourism is not new; Suitable once hosted Beam robots at the base of the Eiffel Tower to encourage remote visits. And the hardware really will, ultimately, become less important than the software. If you think this sounds like a fringe play, consider that a research group at Microsoft lead by Dr. Kori Inkpen recently filed a U.S. patent for “immersive telepresence“, an Uber-like system of matching device-carrying locals with distant users interested in remotely experiencing their area.
“There are various scenarios where a person may want to view a far away locale, but may not be able to do so themselves,” declares the Microsoft patent text in a feat of dry understatement. A tourism-focused telepresence robot—or a camera-carrying human you can boss around—could be the next best thing to being there, if you can handle the lack of direct physical experience. It’s almost retrograde in the face of futuristic promises of virtual reality incarnations of distant places. But even if the worst that can happen is your robot tripping over a carpet, that small degree of messy anarchy makes the remote world more real than a perfect simulation ever could.
This story is part of the Culture Trip Special: Limits collection.