Numéro : 2532 - Year : 2008
The Robosail project, lessons learned
Pieter ADRIAANS, University of Amsterdam
Martijn VAN AARTRIJK:, University of Amsterdam
International Autonomous Surface Symposium (ASSS 2008)
In this paper we describe the lessons learned from the Robosail project.
The project ran from 1997 to 2005, with the aim of building a self-learning, intelligent auto pilot and navigation aid for a single handed sailing yacht. The Robosail yacht participated in a variety of races, with the performance in the Dual Round Britain 2002 (l st in class, 4 th overall) as best result. Spin-offs of the RoboSail project were used on other internationally renowned racing yachts as well. The project consisted of realizing an adaptive system that would assist a single handed sailor to achieve an on average better performance in a race. To this end, we created a hybrid architecture using agent technology, employing machine learning, data mining and rule-based reasoning techniques, presented to the user by an interface targeted to a possibly fatigued and stressed skipper. Apart from describing the system we try to generalize our findings, and argue that sailing is an interesting paradigm for a class of hybrid systems that one could call Skill-based Systems. In these systems human skills are combined with techniques form other domains like operations research, machine learning and data mining. Naturally this leads to issues concerning symbol grounding, modelling of human judgment, hybrid architectures and many other fundamental questions relevant for the construction of applications in this complex domain. We also discuss a number of practical issues, amongst others: legal racing issues, testing techniques and various functional and technical problems we faced during the project.
This paper is written in English
Advanced search
Year