Tutorial on "Mobile Robots"
(Uwe R. Zimmer)
Mobile robots are a motivating and hard test environment for a wide range of methods. The success of each method can be evaluated according to the following criteria, which can be categorised in three fields:
- Realworld abilities, meaning the ability of
handling realworld data, of interacting with the
environment, and of handling dynamic scenes.
- Adaptability or learning abilites of a mobile
robot including subcriteria like spontaneous learning,
stable learning, etc.
- Cognitive complexity or competence, expressed
by the ability of symbolic or qualitative reasoning.
An introduction to principal mobile robot problems together with the definition of some common (partly diverging) goals will be the base for the following discussion of recent projects.
Concepts like Artificial Neural Networks, Fuzzy Logic, Genetic Algorithms, Behaviour Based Approaches, Expert Systems, 'classical' control theory and others are employed in recent mobile robot applications. A couple of projects representing several directions in the mobile robotic research field will be discussed and compared, according to the 'realword', 'adaptability', and 'cognitive competence' criteria.
Currently a number of links to projects discussed in the tutorial is available. If you're interested in further material please send me a short note.
Discussed projects:
Other Publications | Other Talks