Science Autonomy in the Atacama - Robotics Institute Carnegie Mellon University

Science Autonomy in the Atacama

Workshop Paper, ICML '03 Workshop on Machine Learning Technologies for Autonomous Space Applications, August, 2003

Abstract

We have recently embarked on a three-year project, funded by the NASA ASTEP program, to develop robotic astrobiology in the process of learning the limits of life in the Atacama desert of Chile. We see this as an opportunity to develop a more science-aware rover: one that, on encountering a new area, can select interesting features, perform initial experiments, and selectively return relevant data, all before receiving feedback from the science team. Several components of the proposed science autonomy system can make use of classifiers (is this the kind of rock we are looking for?) and clustering algorithms (is this rock like anything we have already sampled?). The unknown character of unexplored areas motivates use of on-line learning techniques.

BibTeX

@workshop{Smith-2003-16891,
author = {Trey Smith},
title = {Science Autonomy in the Atacama},
booktitle = {Proceedings of ICML '03 Workshop on Machine Learning Technologies for Autonomous Space Applications},
year = {2003},
month = {August},
keywords = {space, rover, science, science autonomy},
}