Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
High-speed SLAM for Rail Vehicles
The goal of this work is to develop a high-speed SLAM system that is suitable for usage on rail vehicles (e.g. trains). As main sensor modalities, we want to use a high-speed 1D laser scanner and/or an event-based camera (DVS).
Localization for railway vehicles such as trains is crucial to enable a safe and efficient usage of the infrastructure. Currently, railway safety is provided by infrastructure-side beacons that report the vehicle position and are used for interlocking. In order to make the system more efficient, continuous on-board localization is shown to be highly beneficial.
In this project, we want to leverate the high-speed properties of a 1D laser and an event-based camera (DVS) to map the most relevant parts of the vehicle environment to enable localization and obtain robust and continuous pose estimation even if the vehicle is moving at high speeds (up to 150 km/h).
Localization for railway vehicles such as trains is crucial to enable a safe and efficient usage of the infrastructure. Currently, railway safety is provided by infrastructure-side beacons that report the vehicle position and are used for interlocking. In order to make the system more efficient, continuous on-board localization is shown to be highly beneficial. In this project, we want to leverate the high-speed properties of a 1D laser and an event-based camera (DVS) to map the most relevant parts of the vehicle environment to enable localization and obtain robust and continuous pose estimation even if the vehicle is moving at high speeds (up to 150 km/h).
- Literature review
- Getting familiar with 1D LiDAR and DVS
- SLAM concept for 1D structure
- Map structure
- Map building infrastructure
- Localization pipeline in given map
- Experimental or simulation evaluation of the chosen approach
- Literature review - Getting familiar with 1D LiDAR and DVS - SLAM concept for 1D structure - Map structure - Map building infrastructure - Localization pipeline in given map - Experimental or simulation evaluation of the chosen approach
We are looking for a highly motivated student with experience in C++ programming and interests in computer vision and robotics. Previous knowledge in ROS, Arduino and OpenCV is a plus.
We are looking for a highly motivated student with experience in C++ programming and interests in computer vision and robotics. Previous knowledge in ROS, Arduino and OpenCV is a plus.
If you are interested, please contact Florian Tschopp (florian.tschopp@mavt.ethz.ch), David Hug (david.hug@mavt.ethz.ch) and Andrei Cramariuc (andrei.cramariuc@mavt.ethz.ch), sending your CV, transcript of records and a short paragraph on why you want to work on this project.
If you are interested, please contact Florian Tschopp (florian.tschopp@mavt.ethz.ch), David Hug (david.hug@mavt.ethz.ch) and Andrei Cramariuc (andrei.cramariuc@mavt.ethz.ch), sending your CV, transcript of records and a short paragraph on why you want to work on this project.