Formalization of the traffic world in the C action language

Reasoning about actions and effects of actions is an important task in Artifical Intelligence, with connections to knowledge representation and planning. Many formal methods for representing actions and inferring their effects have been developed over the years (e.g. action languages, fluent calculus, situation calculus). Howewer, the examples formalized so far have been "toy" domains of very small sizes. Successful formalizations of scenarios of nontrivial size are needed in order to show that these methods are suitable for real applications and to assess the strong and weak sides of different methods. The C action language is a logic programming language designed to represent the effects of actions on fluents. In this thesis we formalize the TRAFFIC scenario world - a domain of moderate size, specified at the Logic Modelling Workshop at http://www.ida.liu.se/ext/etai/lmw/ - using the C action language. Example planning problems using the formalization are successfully solved using the Causal Calculator - available at http://www.cs.utexas.edu/users/tag/cc/ -, a program for planning and querying in action domains. The formalization is contrasted with previous work on the TRAFFIC world, namely the formalization of A. Henschel and M. Thielscher using the fluent calculus.