expert meetings

The weeks before easter I scheduled two expert meetings.
The first one was with Emilia Barakova, Ph.D. Faculty of Industrial Design and expert on intelligence in products and human interaction measuring http://www.idemployee.id.tue.nl/e.i.barakova/

I started this meeting with the idea that learning algorithms would be a good solution to make my sailing vessel autonomous. In the discussion with Emilia I found out that learning algorithms aren't always the best solution for a problem. After the meeting I worked with some models based on Q-learning. I found out that the software keeps making the same mistakes over and over again, which is a problem with a real sailing vessel. Also the task the boats needs to perform with the given sensor data is already known if the wind is coming from the north and your heading is also north, the boat needs to tack upwind. This way of programming the autonomous behavior is also much easier.

The second meeting was with René Ahn.
He also works on the faculty of Industrial design in the Design Intelligence group.

In this meeting the focus was on the user. Which information does the user needs to control the boat, and which information does the user needs to give to the sailing vessel to keep it operational.

Another big issue I stumbled upon was communication. How to communicate with your own vessel when it is on the other side of the world in the middle of the ocean. Answers on this topic will come soon hopefully.