ABSTRACT. This tutorial focuses on the representation and semantics of change. The basic language is the situation calculus. This tutorial covers both the well-known logical foundations and the recent developments in the situation calculus. Specifically, the initial part of the tutorial includes logical characterizations of qualitative change within the situation calculus, and the middle part presents a recent approach that extends the previous version of the situation calculus with new representations for reasoning about quantitative change over time. It turns out that this new version of temporal situation calculus can serve as a basis for the development of a new lifted heuristic planner that can solve the planning problems in the mixed discrete-continuous domains. It is ``lifted" in a sense that it works with action schemata, but not with instantiated actions. These hybrid domains include not only the usual discrete actions and qualitative properties, but also continuous processes that can be possibly initiated or terminated by actions. The final part of the tutorial includes an overview of the experimental results collected from running a tentative implementation of a new planner on a selected benchmarks designed by the planning community for the hybrid domains. Since this tutorial focuses on logical representations and semantics, there will be no discussion of the algorithmic or implementation details related to planning. No previous background in the situation calculus or planning is required. However, it is expected that the attendees have basic understanding of knowledge representation and working knowledge of classical first order logic.

See details at  https://www.cs.ryerson.ca/~mes/HTSC-Tutorial-KR2022/

ABSTRACT. This is the 2nd part of the KR-2022 tutorial on July 31st: the 1st part is from 14:00-15:30. See details at  https://www.cs.ryerson.ca/~mes/HTSC-Tutorial-KR2022/