Space Agencies such as NASA and ESA aim at using resources that are available on the moon to support future space missions, for example to produce fuel for deep space missions or as building material for a lunar base. To achieve in-site resource utilization (ISRU), robotic prospecting missions are necessary to identify the locations that are suitable for ISRU missions. To this end, rock samples need to be analyzed for their mineralogy, for example using microscopic images or spectrometer techniques.
To deploy these instruments, the robot that carries them needs to be in a pose close to the rock sample. To maximize the scientific output of a lunar prospecting mission, the robot should reach as many sampling poses as possible in the available mission time. The goal of this thesis is to create a planning algorithm that selects a pose among a set of feasible poses for each given scientific target. The poses should be selected such that they represent the most time-efficient way for the robot to sample all scientific targets.
A special focus is set on testing the approach in a lunar analog mission scenario, i.e. during field tests with our robot ANYmal in a representative environment.
Space Agencies such as NASA and ESA aim at using resources that are available on the moon to support future space missions, for example to produce fuel for deep space missions or as building material for a lunar base. To achieve in-site resource utilization (ISRU), robotic prospecting missions are necessary to identify the locations that are suitable for ISRU missions. To this end, rock samples need to be analyzed for their mineralogy, for example using microscopic images or spectrometer techniques.
To deploy these instruments, the robot that carries them needs to be in a pose close to the rock sample. To maximize the scientific output of a lunar prospecting mission, the robot should reach as many sampling poses as possible in the available mission time. The goal of this thesis is to create a planning algorithm that selects a pose among a set of feasible poses for each given scientific target. The poses should be selected such that they represent the most time-efficient way for the robot to sample all scientific targets.
A special focus is set on testing the approach in a lunar analog mission scenario, i.e. during field tests with our robot ANYmal in a representative environment.
● Review of available path planning algorithms
● Implementation of the selected approach
● Offline testing of the approach on real datasets
● Online testing during fiel tests with our legged robot ANYmal
● Review of available path planning algorithms
● Implementation of the selected approach
● Offline testing of the approach on real datasets
● Online testing during fiel tests with our legged robot ANYmal
● High motivation for the topic and goal-oriented working attitude
● Experience with programming (C++/Python)
● Prior experience with path-planning algorithms is a plus
● Experience with ROS is a plus
● High motivation for the topic and goal-oriented working attitude
● Experience with programming (C++/Python)
● Prior experience with path-planning algorithms is a plus
● Experience with ROS is a plus
Contact via E-Mail, please attach your CV and transcript of recorders.
● Philip Arm parm@ethz.ch
● cc. Jonas Frey jonfrey@ethz.ch
Contact via E-Mail, please attach your CV and transcript of recorders.