Theses and Projects
Please include a transcript of records with all applications!
Please include a transcript of records with all applications!
Master theses:
Learning non-linear robot kinematics based on pose observation
Concentric tube robots consist of multiple pre-bent stacked, hollow tubes that can be translated and rotated (similar to a telescope) to achieve non-linear trajectories. This allows surgeons to access hard-to-reach regions in the body (e.g. behind the eyes, in the cardio-vascular system) and avoid risky tissue (e.g., blood vessels or nerve fibers in the brain).
In this thesis, a kinematic for a CTR robot should be learnt based on provided pose observations of the complete instrument body and executed/tested with an existing robot. As an optional goal, the model may be trained on sparser data (e.g., only the tip position, or the end points of the niti-tubes) and compared against other input data.
In this thesis, a kinematic for a CTR robot should be learnt based on provided pose observations of the complete instrument body and executed/tested with an existing robot. As an optional goal, the model may be trained on sparser data (e.g., only the tip position, or the end points of the niti-tubes) and compared against other input data.
Please note that this thesis will only be available starting Oct. 2024!
Helpful skills:
- Python
- PyTorch/other AI frameworks
- Basic robot control
- Version control with GIT
Contact person:
| Steffen Peikert | Email: steffen.peikert@fau.de |
Integrating medical instruments into CTR
Concentric tube robots consist of multiple pre-bent stacked, hollow tubes that can be translated and rotated (similar to a telescope) to achieve non-linear trajectories. This allows surgeons to access hard-to-reach regions in the body (e.g. behind the eyes, in the cardio-vascular system) and avoid risky tissue (e.g., blood vessels or nerve fibers in the brain).
In this thesis, a selection of existing medical instruments (e.g., gripper, optical fibers for laser, light or camera, suction/fluid delivery, …) should be integrated in a CTR and, if necessary, minituarized. Next, the influence of the integrated instrument is to be tested on the kinematic (i.e., movement reduction, additional bending due to gravity, force when actuating the instrument) of the robot.
In this thesis, a selection of existing medical instruments (e.g., gripper, optical fibers for laser, light or camera, suction/fluid delivery, …) should be integrated in a CTR and, if necessary, minituarized. Next, the influence of the integrated instrument is to be tested on the kinematic (i.e., movement reduction, additional bending due to gravity, force when actuating the instrument) of the robot.
Please note that this thesis will only be available starting Oct. 2024!
Helpful skills:
- (minituarized) Mechanical engineering
- Basic robot kinematics
Contact person:
| Steffen Peikert | Email: steffen.peikert@fau.de |
Research projects:
Shape setting of NiTi-tubes
Nitinol tubes are bought as a straight tube from manufacturers. However, to use them in non-linear robots like CTR, they must be shape-set by clamping them into a mold and heat-treated for ~10min.
In this reserach project, you will help me set up a joule-heating method to shape-set NiTi-tubes to desired bending curvatures.
In this reserach project, you will help me set up a joule-heating method to shape-set NiTi-tubes to desired bending curvatures.
Please note that we will work with strong currents and that the heated Nitinol can reach more than 500°C
Please note that this RP will only be available starting WT24/25!
Helpful skills:
- Electrical engineering (transistors, basic wiring)
- Electric current modelling
# of Students: 1
Contact person:
| Steffen Peikert | Email: steffen.peikert@fau.de |