Perseverance Rover Collision Model for a range of Autonomous Behaviors
Vandi Verma, Justin Huang, Philip Bailey, Joseph Carsten, Douglas Klein
- Year
- 2022
- Citations
- 4
Abstract
The NASA Mars 2020 Perseverance rover landed in Jezero crater on Mars on 18 February 2021. It is a science mission to collect and cache sample cores for possible return to Earth in the future. Robot collision modeling is traditionally used in robotics for hardware safety for manipulation and sampling. The Mars 2020 Rover Collision Model (RCM) optimizes and extends collision checking in innovative ways to provide a range of onboard autonomous capability on a computationally constrained system. It provides an example of the benefit of systems and operations cognizant software design and development of autonomous systems. RCM predictively checks robotic arm, Remote Sensing Mast, and Drill motion and SuperCam laser firing for collisions with rover hardware. It can autonomously generate a terrain model of the robotic arm workspace after a drive without ground in the loop using data from the rovers' stereo cameras. It predicts collisions with terrain hazards for autonomous robotic arm positioning post drive. RCM predictive laser collision checking can be used during autonomous selection of science targets onboard so only collision free targets are used for autonomous science targeting. It provides safe modular access to state for operations, and allows on-board clients a number of options. It is highly configurable and has 10,844 of the total 60,257 parameters on the spacecraft. None of them are required to be modified on a daily basis. RCM was first used on Mars on Sol 11 for SuperCam laser collision checking and on Sol 12 for the first robotic motion, an arm unstow. It has been in regular use since then. As of Sol 210 RCM has performed 33,862 collision checks on Mars without error, reporting collisions where expected. In addition, tens of thousands of collision checks have been performed using in operations ground tools and testbed testing.
Keywords
Related papers
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Artificial intelligence: a modern approach
1995
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991
A new optimizer using particle swarm theory
R.C. Eberhart, James Kennedy
2002