We present a general theory of topological maps whereby sensory input, topological and local metrical information are combined to define the topological maps explaining such information. Topological maps correspond to the minimal models of an axiomatic theory describing the relationships between the different sources of information explained by a map. We use a circumscriptive theory to specify the minimal models associated with this representation.

Topological and metrical methods for representing spatial knowledge have complementary strengths. We present a hybrid extension to the Spatial Semantic Hierarchy that combines their strengths and avoids their weaknesses. Metrical SLAM methods are used to build local maps of small-scale space within the sensory horizon of the agent, while topological methods are used to represent the structure of large-scale space. We describe how a local perceptual map is analyzed to identify a local topology description and is abstracted to a topological place. The mapbuilding method creates a set of topological map hypotheses that are consistent with travel experience. The set of maps is guaranteed under reasonable assumptions to include the correct map. We demonstrate the method on a real environment with multiple nested large-scale loops.

This article reviews map-learning and path-planning strategies within the context of map-based navigation in mobile robots. Concerning map-learning, it distinguishes metric maps from topological maps and describes procedures that help maintain the coherency of these maps. Concerning path-planning, it distinguishes continuous from discretized spaces and describes procedures applicable when the execution of a plan fails. It insists on the need for an integrated conception of such procedures, that must be tightly tailored to the specific robot that is used - notably to the capacities and limitations of its sensory-motor equipment - and to the specific environment that is experienced. A hierarchy of navigation strategies is outlined in the discussion, together with the sort of adaptive capacities each affords to cope with unexpected obstacles or dangers encountered on an animat or robot's way to its goal.

In recent years there has been increasing interest in constructing cognitive vision systems capable of interpreting the high level semantics of dynamic scenes. Purely quantitative approaches to the task of constructing such systems have met with some success. However, qualitative analysis of dynamic scenes has the advantage of allowing easier generalisation of classes of different behaviours and guarding against the propagation of errors caused by uncertainty and noise in the quantitative data. Our aim is to integrate quantitative and qualitative modes of representation and reasoning for the analysis of dynamic scenes. In particular, in this paper we outline an approach for constructing cognitive vision systems using qualitative spatial-temporal representations including prototypical spatial relations and spatio-temporal event descriptors automatically inferred from input data. The overall architecture relies on abduction: the system searches for explanations, phrased in terms of the learned spatio-temporal event descriptors, to account for the video data.

On the basis of a multiple abstraction levels specification process we developed a representation model for environmental robotic knowledge through the definition of a set of ontologies using a multi perspective approach. A general ontological model for typical indoor environments has been first developed, followed by its specialization using an implementation perspective. Finally, actual implementation of the ontology has been obtained via a XML-based markup language, used to build a repository for robotic environmental knowledge.

We present a logic-based framework in which a qualitative spatio-temporal world model is constructed from local surveys i.e. partial spatio-temporal knowledge. Complete space-time histories are obtained through an abductive process driven by continuity constraints and a library of possibly domain specific behaviour patterns. The abduction technique is circumscription, which implements the heuristic that changes should only occur when forced to. We then discuss various additional heuristics to drive the selection of preferred explanations.

Circumscription has been recognized as an important principle for knowledge representation and common-sense reasoning. The need for a circumscriptive formalism that allows for simple yet elegant modular problem representation has led Lifschitz (AIJ, 1995) to introduce nested abnormality theories (NATs) as a tool for modular knowledge representation, tailored for applying circumscription to minimize exceptional circumstances. Abstracting from this particular objective, we propose LCIRC, which is an extension of generic propositional circumscription by allowing propositional combinations and nesting of circumscriptive theories. As shown, NATs are naturally embedded into this language, and are in fact of equal expressive capability. We then analyze the complexity of LCIRC and NATs, and in particular the effect of nesting. The latter is found to be a source of complexity, which climbs the Polynomial Hierarchy as the nesting depth increases and reaches PSPACE-completeness in the general case. We also identify meaningful syntactic fragments of NATs which have lower complexity. In particular, we show that the generalization of Horn circumscription in the NAT framework remains coNP-complete, and that Horn NATs without fixed letters can be efficiently transformed into an equivalent Horn CNF, which implies polynomial solvability of principal reasoning tasks. Finally, we also study extensions of NATs and briefly address the complexity in the first-order case. Our results give insight into the “cost ” of using LCIRC (resp. NATs) as a host language for expressing other formalisms such as action theories, narratives, or spatial theories.

Introduction and Preliminaries For mobile robots, uncertainty is everywhere. Wheels slip. Sensors are a#ected by noise. Obstacles move unpredictably. Truly autonomous robots (and decision-makers or agents in general) must act in ways that are robust to these sorts of failures and unexpected events which we may think of in general as uncertainty. In this chapter, we attempt to meet uncertainty head-on by explicitly modeling it and reasoning about it. We use the term decision theoretic planning to refer to this broad class of planning methods characterized by explicit accounting for uncertainty. We will consider a number of formulations for the problem of planning under uncertainty and present algorithms for planning under these formulations. Uncertainty can take many forms, but for brevity and clarity we will restrict our attention to only two important types: . Prediction uncertainty occurs when the e#ects of actions are not fully predictable. This can be thought of as an uncertain

This paper describes a method for building a cognitive map of a virtual urban environment. Our routines enable virtual humans to map their environment using a realistic model of perception. We based our implementation on a computational framework proposed by Yeap and Jefferies (Yeap & Jefferies 1999) for representing a local environment as a structure called an Absolute Space Representation (ASR). Their algorithms compute and update ASRs from a 2-1/2D 1 sketch of the local environment, and then connect the ASRs together to form a raw cognitive map. Our work extends the framework developed by Yeap and Jefferies in three important ways. First, we implemented the framework in a virtual training environment, the Mission Rehearsal Exercise (Swartout et al. 2001). Second, we describe a method for acquiring a 2-1/2D sketch in a virtual world, a step omitted from their framework, but which is essential for computing an ASR. Third, we extend the ASR algorithm to map regions that are partially visible through exits of the local space. Together, the implementation of the ASR algorithm along with our extensions will be useful in a wide variety of applications involving virtual humans and agents who need to perceive and reason about spatial concepts in urban environments.

This paper describes a method for building a cognitive map of a virtual urban environment. Our routines enable virtual humans to map their environment using a realistic model of perception. We based our implementation on a computational framework proposed by Yeap and Jefferies (Yeap & Jefferies 1999) for representing a local environment as a structure called an Absolute Space Representation (ASR). Their algorithms compute and update ASRs from a 2-1/2D sketch of the local environment, and then connect the ASRs together to form a raw cognitive map. Our work extends the framework developed by Yeap and Jefferies in three important ways. First, we implemented the framework in a virtual training environment, the Mission Rehearsal Exercise (Swartout et al. 2001). Second, we describe a method for acquiring a 21 /2D sketch in a virtual world, a step omitted from their framework, but which is essential for computing an ASR. Third, we extend the ASR algorithm to map regions that are partially visible through exits of the local space. Together, the implementation of the ASR algorithm along with our extensions will be useful in a wide variety of applications involving virtual humans and agents who need to perceive and reason about spatial concepts in urban environments.