Robots in Reception

We have got thoroughly used to phone trees and trawling through automated responses when we phone companies. In the old days dialing “O” usually cut through the trail, but that often doesn’t cut it today. The next routine sorting job that looks set for our robot friends is the “Receptionist.” I can already hear the cries of horror and derision, but robots are already being optimized for the job. (Optimized sounds so much nicer than being trained, don’t you think?)

Information about the establishment, recording your details, and contacting the person that you’re visiting are all straightforward tasks that the computer in any self-respecting robot could handle without difficulty. But a good receptionist does more than that – they make the visitor feel at ease and appear to be interested in them, that is, listening to them and paying attention to what they say.

Here is the challenge, then. How does a robot show that they are paying attention to someone and are therefore interested in what they have to say? Holthaus et al from Bielefeld U have set out to solve that problem.

They chose a rather plain lady-like robot for their experiments with 111 people coming up to her desk. Her responsiveness was programmed on the basis of proximity. “She” would pay more attention to those who were close compared to those who were a long distance away. That all makes sense to me. We all like to stand in line so the guy at the back of the line gets attended to last.

People said that they felt better about her attention to them when she wasn’t giving them her full attention. If she turned at random, not just her head, but her upper body, and came back to them, they felt that she was more aware of their presence and was in charge of what was going on.

This seems to be counterintuitive, but then nobody ever listens to me with consistent rapt attention. They fidget, look around and some will have successfully spotted somewhere to scuttle off to, so, I guess, a robot doing the same thing would make me think it was just like a real person.

  1. P. Holthaus, K. Pitsch and S. Wachsmuth, Int. J. Soc. Robotics, (2011),                          DOI 10.1007/s12369-011-0108-9

Leave a Reply