Learning to Improve Earth Observation Flight Planning

Robert A. Morris, Nikunj Oza, Leslie Keely, Elif Kurklu, Anthony Strawa

This paper describes a method and system for integrating machine learning with planning and data visualization for the management of mobile sensors for Earth science investigations. Data mining identifies discrepancies between previous observations and predictions made by Earth science models. Locations of these discrepancies become interesting targets for future observations. Such targets become goals used by a flight planner to generate the observation activities. The cycle of observation, data analysis and planning is repeated continuously throughout a multi-week Earth science investigation.

Subjects: 12. Machine Learning and Discovery; 1.12 Scheduling

Submitted: Mar 20, 2008


This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.