The VIPER Atacama Mission - 9 July 97

Carlos Ernesto Guestrin, from the desert

Today I went on a panorama hunt, seeking to become more familiar with the region and looking for good places for data acquisition. I decided to spend some time in the region within 20km of the ops-box, which contains a nice mountain range were Nomad's hill-top station is located.

I took three panoramas in an area of about 10km radius in order to test the multiple panoramas algorithm. Instead of estimating the position from just one panorama, this algorithm uses data from several panoramas to nail down the position.

During todays run, I found it difficult to calibrate the compass. What I hear from the team is that there is some source of interference in this region. Although, the Viper system uses the compass reading to determine the orientation of the first image in the panorama, the position estimation algorithm can handle uncertainties in this value.

I also found that the "sensor head", which contains the camera, compass and inclinometers, could not be kept level throughout the 360degrees rotation. I haven't had time to process all of today's data, but this seems to be an extra source of error.

The soil here feels like it was plowed, you sink about 5cm on every step. Driving around was fun! It took the "road" for a while, then went off-road to find a nice place to take images.

The views here are breathtaking. Although, I expected the scenery to be monotonous, I found numerous different landscapes, and they change with lighting conditions. It's very impressive.

Tomorrow, I would like to acquire and process more data. I will also look into the leveling of the sensor head.

  back to Viper system

This work has been conducted at the Robotics Institute at the School of Computer Science, Carnegie Mellon University. It has been partially funded by NASA; Fabio Cozman has a scholarship from CNPq (Brazil). We thank these four organizations for all their support.