in this paper, we develop an account of the kind of deliberation that an agent that is doing planning or executing high-level programs under incomplete information must be able to perform.

Deliberation in highly-dynamic domains such as robotic soccer requires a rich representation language that can deal with continuous change, uncertainty, and multiple agents, among other things.

An effective method is presented for deriving state knowledge in the presence of sensing actions. It is shown how conditional plans can be inferred with the help of a generalized concept of plan skeletons as search heuristics, which allow the planner to introduce conditional branching points by need.

. The Qualification Problem arises for planning agents in realworld environments, where unexpected circumstances may at any time prevent the successful performance of an action. We present a logic programming method to cope with the Qualification Problem in the action programming language Flux, which builds on the Fluent Calculus as a solution to the fundamental Frame Problem. Our system allows to plan under the default assumption that actions succeed as they normally do, and to reason about these assumptions in order to recover from unexpected action failures. 1

When prototyping or developing a system for use in research work, it is often necessary to create an entire system, even if only one part of the system is the focus of the research. Free | open-source software offers a solution to this problem, allowing the creation of cost-effective research platforms, utilising peerreviewed, rapidly-developed code that is easily modified. One form of free | open-source software that is regularly used in research projects is engines from 3d games such as Unreal Tournament (Lewis & Jacobson, 2002). Although the core rendering engine is proprietary, the game engine is able to be freely utilised as a reasonably generic rendering engine and physics simulator, and most of the game code is modifiable.

We investigate the role of open source software in the learning, teaching, utilization, and development of educational robotics and its corresponding core technologies. First, we present a survey of current open source solutions for the LEGO MindStorms Robotics Invention System. Next, we discuss the role of open source software in the development of two undergraduate robotics projects.

Abstract. We present a formal framework where a nonmonotonic formalism (the action description language C+) is used to provide robots with high-level reasoning, such as planning, in the style of cognitive robotics. In particular, we introduce a novel method that bridges the high-level discrete action planning and the lowlevel continuous behavior by trajectory planning. We show the applicability of this framework on two LEGO MINDSTORMS NXT robots, in an action domain that involves concurrent execution of actions that cannot be serialized. 1

A thesis submitted in conformity with the requirements

In AI, the problem of selecting (high-level) actions in dynamic and not completely pre-dictable environments translates into the problem of designing controllers that can map sequence of observations into actions so that certain goals are achieved. One approach to high-level controllers is high-level programming. Basically, we imag-ine a system executing a high-level program with respect to a background theory of action. Whereas the background theory describes the characteristics of the domain (ac-tion preconditions, action effects and non-effects, etc.), the high-level program provides strong, but usually incomplete, clues about what the desired sequence of actions should be like. The work presented here combines two recent, but unexplored ideas: Guarded Action Theories and Incremental Program Execution. The result is a new high-level program-ming language, which we call IndiGolog, whose programs, compared with previous lan-guages, are executed in a more practical way with respect to more open-world theories. We provide a theoretical exploration of both guarded theories and program execution, and develop a Prolog implementation for IndiGolog.

Abstract IndiGolog is a programming language for autonomous agents that sense their environment and do planning as they operate. Instead of classical planning, it supports high-level program execution. The programmer provides a high-level nondeterministic program involving domain-specific actions and tests to perform the agent’s tasks. The IndiGolog interpreter then reasons about the preconditions and effects of the actions in the program to find a legal terminating execution. To support this, the programmer provides a declarative specification of the domain (i.e., primitive actions, preconditions and effects, what is known about the initial state) in the situation calculus. The programmer can control the amount of nondeterminism in the program and how much of it is searched over. The language is rich and supports concurrent programming. Programs are executed online together with sensing the environment and monitoring for events, thus supporting the development of reactive agents. We discuss the language, its implementation, and applications that have been realized with it.