Last week we told you about the possibility of robots moving into low-level service jobs, and today brings research news that shows some baby steps in that direction.  UC Davis researchers are creating “follower robots”.  These bots are programmed to track and follow human leaders.  By analyzing the human leader’s movements, the robot effectively predicts where the human will go next, thereby preventing itself from crashing into walls or losing the leader.

Previous models of follower bot were equipped with a camera to track the human’s path, but it became difficult to keep up when the human turned or zig-zagged.  The current robots are outfitted with both a camera and a behavioral cue sensor to note human movements that work as predictive aids.  For instance, if a human points in a direction, it’s likely he’ll move that way.  Humans typically rotate their head 25 degrees about 200 milliseconds before making a turn.  By analyzing slight movements like these, the robot can more accurately predict the leader’s next action. The robots can also pick up on emotional cues that may affect movement, like the way sadness is tied to slower movement and stress or excitement to quicker movement.

As this research progresses, the robots’ functionality could expand to things outside of simple following.  While follwer robots have positive implications for the future of robotic workers, they are quite ominous to those of us who enjoy strolling down unlit city streets at night, occasionally glancing over our shoulders just to make sure we’re not being followed by a heartless, bloodthirsty machine.

[Popular Science]