The Mars Exploration Rovers (MER) have been operating on Mars for more than three years. The extremely high reliability demonstrated by these rovers is a great success story in robotic design. This reliability comes at a high cost, however, both in the initial cost of developing the rovers and in the ongoing operational costs for their mission extensions. If it were possible to design rovers with reliability more in line with their mission requirements (in the case of MER, 90 days), considerable cost reductions could be achieved. This will be even more important for future planetary robotic missions due to greatly increased mission durations.
In this paper we present an overview of our ongoing research in the area of predicting robot mission reliability, and we show how a mission designer can trade off reliability against costs in order to find an optimal reliability target for a given robotic mission. Our results show that for a given mission there is an optimal reliability range with respect to cost and that having rovers with reliability that is too low or too high is suboptimal from an economic standpoint. This suggests that a better cost-reliability tradeoff can be obtained by "planning to fail" by designing rovers which have lower reliability than current legacy designs.
|planetary rovers, mission design, mission cost, reliability, failure, risk|
Associated Center(s) / Consortia:
Vision and Autonomous Systems Center
Associated Project(s): Reliability of Mobile Robot Teams
Number of pages: 5
|Stephen B. Stancliff, John M. Dolan, and A. Trebi-Ollennu, "Planning to Fail - Reliability as a Design Parameter for Planetary Rover Missions," Proceedings of the 2007 Workshop on Measuring Performance and Intelligence of Intelligent Systems (PerMIS '07), August, 2007, pp. 218 - 222.|
author = "Stephen B Stancliff and John M Dolan and A. Trebi-Ollennu",
title = "Planning to Fail - Reliability as a Design Parameter for Planetary Rover Missions",
booktitle = "Proceedings of the 2007 Workshop on Measuring Performance and Intelligence of Intelligent Systems (PerMIS '07)",
pages = "218 - 222",
month = "August",
year = "2007",
|The Robotics Institute is part of the School of Computer Science, Carnegie Mellon University.|
Contact Us | Update Instructions