Master Thesis: “Bringing autonomy to our surface water drone – autonomous sailing between Arnhem and Wageningen”
Currently, our unmanned surface vessel, the C-Drone, can automatically sail a track based on a set of given waypoints. We want to use the C-Drone to autonomously perform missions such as mapping current profiles and water quality for prolonged times without supervision. We are looking for a student that can help us realise this.
With the increased availability of computation power and sensors, the use of autonomous vessels for inspection tasks has risen significantly. However, most of the missions consist of sailing a track without colliding with obstacles. Our idea is that the real use of autonomy for relative small surface vessels will be if a mission can be specified on a high abstraction level, and not broken down to detailed tasks by the user. However, the final vehicle control system needs to be provided with settings for a velocity and heading.
In this assignment we want to break down the high level mission automatically to a set of sequential/parallel tasks such that our C-Drone can execute it and really becomes autonomous. These tasks might change during the execution of the mission due to environmental influences. A mission is tested on the river with the C-Drone to validate the approach.
The foreseen elements that needs to be tackled in this assignment when the mission is the mapping of the current profile of the middle of the Rhine between Wageningen and Arnhem:
- define a (recursive) framework in which the mission can be decomposed in smaller tasks that need to run sequential or in parallel. The framework needs to be suitable for a single vessel, or for a swarm of vessels. In the case of measuring the current the decomposition can be to run the tasks ‘check own status’, ‘go to middle of the river’, ‘sail to Arnhem’, ’repeatedly take measures’, ‘return to harbour’, while it always needs to avoid collisions and check if it has enough energy to return back
- a literature study on different approaches that are currently be used for this
- implement the framework that can interface with the vehicle control software in the C-Drone
- implement common tasks such as anti-collision, path-planning (preferable with open-source software)
- validate the approach by executing the mission.
Although the platform for the assignment will be the C-Drone, the approach should hold for the other vehicles with which we investigate autonomy: a 6m RHIB, and an underwater vehicle. This requires that the framework is generic. Furthermore, the goal is that the mission is actually executed. Practical issues might need to be solved for this.
The duration will be between 6 and 9 months. The project can start at September 1st.
The student is encouraged to give the assignment a personal spin. We are looking for a student working on its Master’s thesis.
Department and supervisor
You will execute your internship at the Performance at Sea department of MARIN. For more information please contact Bas de Kruif (B.J.d.Kruif@marin.nl)
Solliciteer / Apply