Telepresence robots are deputising for their human controllers in the workplace, but can they really fit in?
YOU wander around the office, stopping off in the kitchen to chat with colleagues before heading down the hallway to an impromptu design meeting. Afterwards, you head to a colleague’s office to continue the discussion.
It is a fairly typical day at work – except that you are in London, while all the people you have been talking to are in your company’s New York office. You have, in effect, beamed yourself across the Atlantic in the guise of a telepresence robot.
Advertisement
These robots are essentially a video camera, speaker and screen on wheels, and they can be controlled from anywhere in the world using a web browser. They allow managers to keep an eye on their factories overseas and multinational teams to collaborate on projects – all without setting foot on a plane.
As the robots begin to appear in more and more offices and factories, they are poised to transform the way we work and interact with our colleagues. For one thing, the robots’ mobility lets people in remote locations talk to colleagues without resorting to the videoconference facilities of the meeting room, says Leila Takayama of , a robotics company in Menlo Park, California. “Telepresence is really for informal communications, for the gatherings that happen before and after the big meetings, where real decisions are made and real team-building gets done,” she says.
Another California-based firm, , recently began selling its QB telepresence robot as an alternative to conventional videoconferencing systems. Similar robots are marketed by in Nashua, New Hampshire, whose clients include hospitals and universities. Teenager Lyndon Baty from Knox City in Texas, who suffers from polycystic kidney disease, manages to attend classes at his school using a VGo robot.
At Google, Johnny Chung Lee has built his own telepresent robot so that he can interact with his fiancée when they are apart. Lee, who previously helped develop the Kinect gaming system while at Microsoft, used off-the-shelf components for the robot and wrote a simple application to control it from a built-in laptop.
But can these robots blur the lines between human and machine? In a research project to be presented in May at a , Canada, Takayama and Min Kyung Lee at Carnegie Mellon University in Pittsburgh, Pennsylvania, monitored workers at three companies who were using Willow Garage’s prototype telepresence system.
They found that remote workers using the robots were interacting with colleagues as if they were all physically together. They used the bots to attend meetings and informal gatherings in the office kitchen or games room. What’s more, people liked interacting with the robots.
“Remote workers who were present as robots were interacting with colleagues as if physically together”
“One participant who worked remotely said he had previously felt very invisible, as no one showed an interest in interacting with him,” says Lee. “But once he started using the robot, everyone wanted to talk to him.”
Remote workers can feel a closer connection to their office colleagues than is possible with emails or phone calls, says Katherine Tsui of the University of Massachusetts Lowell, who has also been studying the impact telepresence robots have in the workplace.
“Overall we found that in more ad-hoc situations, and for people who wanted to be closer to their teams, telepresence robots did provide benefits,” she says. “They help to foster a closer connection.”
But a closer connection with what, exactly? Both Tsui and Takayama found that people were confused as to whether they should treat the robot as a machine or a real person.
In Tsui’s study, some workers reported that the robot felt more like a child or pet than a colleague. Takayama found some workers even treated the robot as if it were furniture, using it as a footstool or a post to lean against.
“You wouldn’t put your arm on a human colleague’s head when you were talking to them, but some people will do that to the robot,” says Takayama.
Others treated the robot as if it really were the human controller or “pilot”. They considered it impolite to push the robot to make it go faster or to adjust the volume setting without asking the pilot’s permission.
As for the pilots, some were offended when people made physical contact with their robots, Takayama says. “When pilots felt that someone was invading their personal space, they would say things like, ‘He was up in my face, he shouldn’t have been that close’, or ‘I wish they wouldn’t poke it there’. It sounded to me like the way a person would respond to their physical body’s space being invaded.”
This is in stark contrast to conventional videoconferencing, when you wouldn’t expect someone to object to the camera being touched. “That is the biggest difference between this and other conferencing systems – the person’s sense of self seems to be somehow incorporating the technology as if they are the machine, and [they feel that] other people should respect that,” says Takayama.
Direction unknown
If telepresence robots are to be smoothly rolled out into our workplaces, their developers will have to find a way to stop them crashing into walls.
Google’s Chris Uhlik has tested both the VGo and Anybots QB robots at the company’s offices in Mountain View, California. He found that fluctuations in the speed of the network connection meant the video feed would sometimes cut out, leaving the controller steering blind. “You can be driving down the hall and the robot will simply give up for a few minutes, and you end up driving into a wall,” he says. Forcing the system to update the video as best it can despite a poor connection could help get around the problem, he says.
