TU Darmstadt / ULB / TUbiblio

Autonomous Mission Planning And Execution For Two Collaborative Mars Rovers

Victoria, Juan Delfa and Yeomans, Brian and Gao, Yang and Stryk, Oskar von (2015):
Autonomous Mission Planning And Execution For Two Collaborative Mars Rovers.
In: ASTRA, [Conference or Workshop Item]

Abstract

One of the most important lessons learnt from more than ten years of robotic exploration in Mars is the high value of autonomy in terms of operations and science return. So far, only a few full-scale experiments to foster autonomy for planetary exploration have been conducted in Europe. One of the most relevant is the recently completed FASTER FP7 framework project, with one of the main scientific contributions being the application of autonomous planning and execution under uncertainty for two collaborative rovers in a Mars scenario where it is not possible to have humans in the loop. This paper presents the results of implementing the QuijoteExpress (abbreviated QE) mission planner and SanchoExpress executive to tackle these problems. QuijoteExpress is a novel platform-independent planner which strengths are performance, plan robustness and expressiveness. QuijoteExpress was able to produce robust collaborative plans for both rovers with only minor tuning under undeterministic and unstructured scenarios. SanchoExpress is a ROS-based, platform-independent timeline executive that can command multiple subsystems in parallel. SanchoExpress (abbreviated SE) can encode platform-dependent knowledge including repair methods for some errors which allow the executive to fix the plan in certain fault scenarios without the need of the replanner, therefore contributing to a more robust execution of the plan. The combined system was found to operate successfully and reliably during the FASTER field trials, enhancing the systems autonomy and enabling complex behaviour to be executed in a robust and fault tolerant manner.

Item Type: Conference or Workshop Item
Erschienen: 2015
Creators: Victoria, Juan Delfa and Yeomans, Brian and Gao, Yang and Stryk, Oskar von
Title: Autonomous Mission Planning And Execution For Two Collaborative Mars Rovers
Language: German
Abstract:

One of the most important lessons learnt from more than ten years of robotic exploration in Mars is the high value of autonomy in terms of operations and science return. So far, only a few full-scale experiments to foster autonomy for planetary exploration have been conducted in Europe. One of the most relevant is the recently completed FASTER FP7 framework project, with one of the main scientific contributions being the application of autonomous planning and execution under uncertainty for two collaborative rovers in a Mars scenario where it is not possible to have humans in the loop. This paper presents the results of implementing the QuijoteExpress (abbreviated QE) mission planner and SanchoExpress executive to tackle these problems. QuijoteExpress is a novel platform-independent planner which strengths are performance, plan robustness and expressiveness. QuijoteExpress was able to produce robust collaborative plans for both rovers with only minor tuning under undeterministic and unstructured scenarios. SanchoExpress is a ROS-based, platform-independent timeline executive that can command multiple subsystems in parallel. SanchoExpress (abbreviated SE) can encode platform-dependent knowledge including repair methods for some errors which allow the executive to fix the plan in certain fault scenarios without the need of the replanner, therefore contributing to a more robust execution of the plan. The combined system was found to operate successfully and reliably during the FASTER field trials, enhancing the systems autonomy and enabling complex behaviour to be executed in a robust and fault tolerant manner.

Title of Book: ASTRA
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Simulation, Systems Optimization and Robotics Group
Date Deposited: 20 Jun 2016 23:26
Identification Number: delfa2015astra
Export:
Suche nach Titel in: TUfind oder in Google

Optionen (nur für Redakteure)

View Item View Item