Sincere thanks for getting the point of the story (and for not expecting all coverage of robots and AI to be wild speculation about the next century or millennium).

It's definitely not much more complex than that. The only technical innovation is the use of computer vision to recognize that a door has opened. But this thing is designed to operate almost continuously, and probably costs exponentially less to build. Those experimental bots almost always insanely expensive, and constantly breaking down.

Do you encounter any robots that do that? iRobot's AVA, or even those bots in hospitals, will stop and wait for a clear path, but it would take some really bad engineering, and a deep urge to be sued, to build a robot that plays chicken with pedestrians, This bot specifically weaves around obstacles, instead of coming to a dead stop like a big dummy.

Really sorry about that. Letting my editors know right now, but has that been the case for a while?

One of the things I didn't include in the story, since it was more of a hunch on the part of the robot's makers than anything based on the pilot, is the notion that people might be more likely to make service item requests if they don't have to deal with a human. That could be because they don't want to worry about tipping someone for bringing an item, like a toothbrush, that's ostensibly free, or because they simply don't want to deal with someone after a long trip. Or, and this is where some data would be valuable, there's the anecdotal "evidence" of female hotel guests just not being comfortable with opening the door to a stranger, if it's not absolutely essential. Again, not exactly a concrete benefit, but I think it makes a good deal of sense.

This sort of bot isn't sophisticated enough to replace a human runner.