









 |
    
|
Log Book for April 27, 2004
EVA Robotic Assistant Report
Crew 29 Reporting
- Today is the first attempt at the Sol 1 activitiy of the Pooh's Corner scenario.
Sol 1 requires a lengthy traverse by Boudreaux to collect science data from four locations identified earlier as potentially high-value scientific targets.
In this scenario, the science data consists of a series of high-quality color panorama images vitally important for the Remote Science Teams to plan the Sol 2 activity for two astro-geologists and their robot assistant Boudreaux.
- From 9-11am, the ERA team prepared for the Sol 1 activity. A fresh software load was distributed to the onboard computers, fully-charged batteries were installed, systems were checked, and the robot maneuvered to the designated starting waypoint called "Hab" located between the MDRS Habitat and the ERA's camp.
When human-robot teams do go back to the Moon or to Mars, software reloads will consist of small bug-fixes, since flight qualified systems are more rigourously tested than the research-grade software onboard Boudreaux today.
- Nathan and Rob prepared for their roles as "RoboChase"-- to follow the robot, ensure crew safety (including Boudreaux), and report any anomalous behavior back to the Hab. Nathan and Rob put on hats and sunscreen, and carried radios and water bottles.
For a real Mars mission, the role of RoboChase will not exist, nor should it be necessary. The EVA Robotic Assistant of the future will be a fully capable and trusted teammate.
- Kim and Jeff moved into mission control on the 2nd floor of the Habitat. They setup remote workstations for monitoring Boudreaux in every aspect: communications, vehicle health, onboard processing.
One workstation was dedicated to projecting live video from Boudreaux's onboard cameras onto a large screen to provide Mission Control with additional situational awareness during the run. Boudreaux has two onboard digital firewire cameras mounted about 5' above ground on pan-tilt mechanisms, one on the front of the robot and one on the rear. Both cameras can be controlled from HabCom to look around, zoom, and snap photos. These same cameras are used for taking the panorama images.
- Kim and Jeff started Boudreaux's various software processes and Boudreaux received it's first command from the Mobile Agents software at 17:58 UTC - a command to autonomously move, with obstacle avoidance and path planning, from it's current location to a designated GPS waypoint approximately 40 meters ahead.
- The first software upset occurred at 18:17 UTC, about 19 minutes into the run.
- The robot traveled a total distance of 1162 meters today, a new distance record for Boudreaux, and a rare event for mobile autonomous robots working in such extreme environments. Approximately 15 meters of the total distance were under joystick control by RoboChase to deal with upset events in the robot systems.
- Overall, the run lasted from 17:58 - 00:42 UTC for an elapsed time of 6.73 hours, a new endurance record for Boudreaux. One battery swap for the onboard laptop computers and one battery swap for the drive motors were required.
A perfect run would have been considerably shorter in duration. Several "holds" were necessary during the run to deal for the ERA team to deal with upsets and events.
- During the run, the following robot events occurred:
| Hardware |
Quantity |
Description |
| drive electronics |
4 |
multiple occurances of: PIC read failures (new failure class, under investigation) |
| GPS |
1 |
LOS from GPS basestation correction radios (corrected) |
| Camera |
3 |
multiple occurances of: loss of both video feeds (cause under investigation) |
TOTAL
|
8
|
|
| Software |
Quantity |
Description |
| path planner |
4 |
multiple occurances of: spontaneous aborts |
| obstacle avoidance |
6 |
multiple occurances of: point turns being commanded when disallowed (fix inwork) |
| executor |
1 |
underspecified goal waypoint data (fixed) |
| panorama server |
1 |
cardinal compass directions incorrect in image overlay (fixed) |
| obstacle map |
multiple occurances of |
switching map sections causes temporary erratic base motion "the crazy Boudreaux" (fix inwork) |
| TOTAL |
12 |
|
- In terms of Sol 1 scenario goals, the robot (a) successfully traversed (albeit slowly) to
three of the four designated locations, (b) successfully created panoramas at the first two (the Sol 1 run was aborted at the third location due to MDRS schedule constraints and a software issue in the Brahms Mobile Agents software that failed to command Boudreaux to take a panorama there, and (c) provided situational awareness to HabCom in the form of live video.
|
|
|