Working/Master Student (f/m/x): LiDAR-Based Satellite Pose Estimation for Orbital Robotic Operation
Background:
There is a growing demand for satellite servicing and debris removal from orbit in order to ensure a continuation of orbital space flight. In particular, life extension of expensive space hardware as well as reducing the risk of collision with failed satellites and components is necessary for maintaining and further developing the significant orbital infrastructure that we rely on today. At the Institute of Robotics and Mechatronics we investigate robotic solutions to the required space operations. Among the main challenges is the pose estimation and tracking of an uncooperative target satellite based on observations from the orbital servicer platform, which is a prerequisite of autonomous robotic manipulation. We have developed an initial solution that makes use of a pair of LiDARs mounted on the servicer. Based on this previous work, some further investigation and extensions are desirable, which we offer as a project to a master student.
Your tasks:
- Generation of LiDAR datasets from simulation: different target satellite shapes in different poses relative to a LiDAR system mounted on a servicing satellite
- Based on our existing neural network architecture: investigation of ensemble methods for pose estimation and uncertainty quantification
- Investigation of cases with satellite pose ambiguity
- Investigation of possible improvements to our existing neural network architecture
Your qualifications:
- Final year master student in computer science, aerospace engineering, robotics, or related
- Strong background in machine learning, in particular deep learning
- Knowledge of, preferably even experience with methods for 3D point cloud processing
- Knowledge of, preferably even experience with methods for object pose estimation
- Proficiency in Python programming, in particular using TensorFlow or PyTorch
- Experience in C++ programming
- Experience in using MATLAB or Mathematica
Start at the earliest possible date. Contract duration is 8 months (2 months as working student, followed by 6 months for the thesis project).
For application, first make contact (see below) to indicate your interest. We will then get back to you.