Miller Robotic Interface II Manual de usuario descargar pdf (Pagina 39)

hypothesized that having both the color changing, as well as how many bars were filled in,

would allow for a better understanding of the robot’s situation. Having the ranging information

being displayed as numbers was thought to be too hard to track and mentally exhausting because

the user would have to be the one processing what the numbers meant.

The rear view camera was moved to the top right of the main video screen. We chose to move

it here so that it would mimic the placement of the rear view mirror in a car. An automobile

driver needs to look up and to the right to see their rear view mirror; likewise, our users need to

look up and right to see the rear view camera.

Directly above the main video screen is the first implementation of the suggestion system that

was previously mentioned in section 4.4. It is recommending that the lights be turned off to

conserve battery power, as well as a switch to safe mode In the mode bar, safe mode is also

being highlighted, indicating that it is being suggested, but the background is still red and the text

says “TeleOp Mode” informing the user that the robot is in teleop mode.

This version of the interface was subject to two studies. The first was a small study consisting of

three participants. This study was conducted to get a general sense if we were on the right track

with the interface and our hypotheses. The second study was conducted in conjunction with

Swarthmore College. On our end, we set out to see if having two opposite facing cameras, one

facing forward and one facing backward, would lead to better SA. Swarthmore was tasked with

seeing if one forward facing camera and one overhead camera, which provided a view of the

robot’s chassis in the video image, would lead to better SA.

5.1.3 Results of the First Study

We began this study with the idea of evaluating our progress to this point by comparing it with

the original INL system [Bruemmer et al., 2004]. The INL system won the AAAI robot rescue

competition that year and was considered to be the best and most complete research USAR

system at the time. The idea was simply to conduct the same user test as was done using the

INL system as explained by Yanco and Drury [2004] and compare the results. In the interest of

full disclosure, however, our system differed in more than just the interface. Both robot

platforms were ATRV-Jr research robots, but our robot had nine more sonar sensors than the

1 2 ... 34 35 36 37 38 39 40 41 42 43 44 ... 79 80

Sin comentarios

Miller Robotic Interface II Manual de usuario Pagina 39