This project comming from the KU Leuven in cooperation with Waterwegen en Zeekanalen N.V. and is designed by 2 students from the faculty industrial engineering as a pilot project. The goal of this project is the creation of an autonomous boat that is capable to traverse the inland waterways found in Europe.
students:
- Natan Biesmans: natan.biesmans@student.kuleuven.be
- Senne Vanstraelen
The system reads in the sensor data and translates it to GPS data. The sensor data is also used as a source in the creation of virtual obstacles creating perimeters surounding the obstacles and hence creating safety perimeters. All these data-points are read in by ROS via a custom implementation of websockets based upon the intelligen tenergy websocket library. In ROS these points are used to create a path to the desired destination. The data from the pathfinder being fed back from ROS is used to determine a heading based upon the current location and the location fed by the pathfinder. The pathfinder also gives the next turnging point of the path to determine the angle of the turn. This is used to slow down the boat when approaching a turn. This slowing down is proportional to the distance and angle of the turn. The outputs are controlleled by PID-controllers that compare the desired headings and speeds with the real headings and speeds to determine the needed output power. This is done in a vectorised way.
The outputs:
- waterjets as bow-propulsion
- azipod
- propeller
Sensor:
- LIDAR: SICK lms511Lite
- GPS: Navigeo ¨Pro
- Ultrasonic Sensors: cluster connected to arduino ethernets
Computing:
- ODROID XU-4: handles ROS
- CRIO 9033: handles LABVIEW
The paths are read in via a off-site database that holds both the paths and constraints used by the autonomous system. Constraints are parameters depecting rules of the waterways used by the boat. These can range from speed limits to detours. These constraints and paths can be created and manipulated by logistic sofware are also written in labview.
At the moment the systems are still in beta with still some bugs in the PID-controllers, due to lags in the system it stays unstable. There several solutions that are currently being investigated, but due to project timings these have to wait till June to be implemented.
Hi guys, this sounds like an AWESOME project. Unfortunately, it is missing some vital details (including who YOU are, which UNIVERSITY you are from, and an email address for us to contact you on if you are successful in becoming an award finalist).
Also, extra details reguarding which NI tools you used, and the overall success of the project would also be very welcome.I would suggest having a look at a couple of the other entries to get a feel for the format:
Competition judging begins soon, so recommend making the admentments quickly too.
Thank you for your interest in the contest. And congratulations on a fantastic project.