Social Robotic Walker
We've had an extra week to polish our Social Robotic Walker project from Interface class. One of the problems with our original presentation was the disconnect between the moving arrow and the static viewer. The arrow is supposed to respond to the movements and orientation of the walker, but that's hard to evaluate when the interface is tied to a desktop computer.Our solution was to create a mobile prototype.
We loaded the interface onto Andy's Powerbook and secured it to the top of a box of paper. Then we strapped the whole thing to the seat of a rolling chair. Using a wireless keyboard, we were able to make the interface respond like a touchscreen, complete with an arrow that moves as if it were being controlled by GPS. We replaced our nursing home photos with some pictures that Dan took of the rooms down the hall from the graduate studio and ended up with a testable prototype [Flash 1MB]. To see how it worked, click one of the locations, then use the arrow keys on your keyboard to fake moving the arrow around.
In order to evaluate our new interface, we recruited some people who didn't know their way around Margaret Morrison hall and sent them off [Quicktime 1.4MB] with the "walker," using the arrows for navigation.
We've made some subtle alterations in light of our findings. For instance, it turned out that people simply don't make the connection that the arrow is an invitation to move forward. Once they understand the rationale, they can easily follow the arrow's directions, but that initiation is a major hurdle.
Posted November 13, 2003
CMU Grad Studio