Miller Robotic Interface II Manual de usuario descargar pdf (Pagina 12)

1.2 Contributions

This work describes an interface design that is a result of an evolutionary process. This design

was validated through user testing, which showed improved awareness of the robot’s

surroundings with each new version.

Few researchers in the HRI domain are iterating interface designs via user testing. Notable

exceptions are the Idaho Nation Laboratory (INL) and also Swarthmore College. Due to the

relative paucity of literature describing HRI design iterations, one of this thesis’ contributions is

the documentation of our evolutionary process.

This work puts into action many of the guidelines produced by Yanco, Drury and Scholtz [2004]

and Scholtz et al. [2004]. We provide a map of where the robot has been as well as fused sensor

information. We do not have the user tab through multiple windows to find the information

they need. All of the important information is displayed in a single window. We provide more

spatial information about the robot in the environment. This spatial information takes the form

of a map, as well as displaying the current distance sensor readings in an easy to interpret

distance panel. This information makes it easy to know if the robot is close to an obstacle or

not. We also provide information on which camera is currently the main one and we use a

crosshair overlaid on the video to indicate the current pan/tilt position of it.

Some original guidelines were also created as result of this work. These guidelines can be added

to those already mentioned from prior work. For instance, using this system, we proved that

having multiple cameras, especially one facing the rear of the robot, greatly improved situation

awareness. Therefore, we state that to improve SA most effectively, if at least two cameras are

present on the robot system, one should face forward and one backward. We also show that

having the ability to see at least part of the robot’s chassis in the video stream also leads to

improved SA which follows along the guideline about being able to use the cameras to inspect

the robot. Also, for our interface, we presented all of the important information on or around

the video screen. Operators pay attention primarily to the video, so information will only be

noticed if it is overlaid on the video, or directly adjacent to it. Although we did not explicitly test

the validity of this claim in our experiments, it has been seen that users ignore information not

near the video screen. Therefore, as a guideline, we state that all important information needed

1 2 ... 7 8 9 10 11 12 13 14 15 16 17 ... 79 80

Sin comentarios

Miller Robotic Interface II Manual de usuario Pagina 12