Andrew Rock

EFFICIENT DEFEASIBLE REASONING SYSTEMS

For many years, the non-montonic reasoning community has focussed on highly expressive logics. Such logics have turned out to be computationally expensive, and have given little support to the practical use of non-monotonic reasoning. In this work we discuss defeasible logic, a less-expressive but more efficient non-monotonic logic. We report on two new implemented systems for defeasible logic: a query answering system employing a backward-chaining approach, and a forward-chaining implementation that computes all conclusions. Our experimental evaluation demonstrates that the systems can deal with large theories (up to hundreds of thousands of rules). We show that defeasible logic has linear complexity, which contrasts markedly with most other non-monotonic logics and helps to explain the impressive experimental results. We believe that defeasible logic, with its efficiency and simplicity, is a good candidate to be used as a modeling language for practical applications, including modelling of regulations and business rules.

Propositional plausible logic: Introduction and implementation

Plausible Logic allows defeasible deduction with arbitrary propositions, and yet when sufficiently simplified it is very similar to the Defeasible Logics of Billington and Nute. This paper presents Plausible Logic, explains some of the ideas behind the definitions, applies Plausible Logic to an example, and proves a coherence result which indicates that Plausible Logic is well behaved. We also report the first complete implementation of propositional Plausible Logic. The implementation has a web interface which makes it available to researchers and students everywhere. The implementation is evaluated experimentally, and is shown to be capable of handling tens of thousands of rules and sufficiently many disjunctions for realistic problems.

Efficient defeasible reasoning systems

For many years, the non-monotonic reasoning community has focussed on highly expressive logics. Such logics have turned out to be computationally expensive, and have given little support to the practical use of non-monotonic reasoning. In this work we discuss defeasible logic, a less-expressive but more efficient non-monotonic logic. We report on two new implemented systems for defeasible logic: a query answering system employing a backward chaining approach, and a forward-chaining implementation that computes all conclusions. Our experimental evaluation demonstrates that the systems can deal with large theories (up to hundreds of thousands of rules). We show that defeasible logic has linear complexity, which contrasts markedly with most other non-monotonic logics and helps to explain the impressive experimental results. We believe that defeasible logic, with its efficiency and simplicity is a good candidate to be used as a modelling language for practical applications, including modelling of regulations and business rules.

An implementation of propositional plausible logic

We report the first complete implementation of propositional plausible logic. Plausible logic is an extension of defeasible logic that overcomes the latters inability to represent or prove disjunctions. This advantage is significant in dealing with practical applications such as the modeling of regulations. The system has a Web interface, which makes it available to researchers and students everywhere. The implementation language chosen was Haskell and some advantages and consequences of this choice are discussed.

Architecture for Hybrid Robotic Behavior

Software architectures for agent technology and robots have been polarized between reactive architectures and architectures based on planning and reasoning. Although hybrid architectures have been shown to offer benefits from both, these seem complicated to integrate. In this paper we integrate the reactive nature of finite state machines and the reasoning capabilities of non-monotonic logics to produce intelligent autonomous robots. In particular, we demonstrate this with a robotic poker player. The robotic player integrates vision, sound recognition, motion control and the reasoning to perform competitively as a player in a complex game with incomplete information.

Is there resonance enhancement of the cross section for vibrational relaxation induced by very low energy collisions? The I2–He system revisited

Rate coefficients for state‐to‐field vibrational relaxation of I2 (B 3Π0u+,  υ’=16) induced by collisions with He at translational temperatures of 2–12 K are measured using state‐selected time‐resolved dispersed fluorescence spectroscopy in a supersonic free jet expansion. These vibrational relaxation rate coefficients in the 2–12 K regime are found to be an order of magnitude smaller than comparable rate coefficients at 300 K. The rate data are compared with calculated classical I2–He encounter rates and with rates computed using quantum mechanical cross sections for I2–He calculated by Schwenke and Truhlar. The role of scattering resonances in determining vibrational relaxation rates at low translational temperatures is discussed.

Using Temporal Consistency to Improve Robot Localisation

Symbolic reasoning has rarely been applied to filter sensor information; and for data fusion, probabilistic models are favoured over reasoning with logic models. However, we show that in the fast dynamic environment of robotic soccer, Plausible Logic can be used effectively to deploy non-monotonic reasoning. We show this is also possible within the frame rate of vision in the (not so powerful) hardware of the AIBO ERS-7 used in the legged league. The non-monotonic reasoning with Plausible Logic not only has algorithmic completion guarantees but we show that it effectively filters the visual input for improved robot localisation. Moreover, we show that reasoning using Plausible Logic is not restricted to the traditional value domain of discerning about objects in one frame. We present a model to draw conclusions over consecutive frames and illustrate that adding temporal rules can further enhance the reliability of localisation.

Modelling behaviour requirements for automatic interpretation, simulation and deployment

In this paper we propose a high level approach to capture the behaviour of an autonomous robotic or embedded system. Using requirements engineering, we construct models of the behaviour where system activities are captured mainly by collaborating state machines while the domain knowledge is captured by a non-monotonic logic. We explain our infrastructure that enables interpretation, simulation, automatic deployment, and testing of the models, minimising the need for developers to code. The approach also minimises faults introduced in the software development cycle and ensures a large part of the software is independent of the particular robotic platform.

Requirements Engineering via Non-monotonic Logics and State Diagrams

We propose to model the behaviour of embedded systems by finite state machines whose transitions are modelled by predicates of non-monotonic logics. We argue that this enables modelling the behaviour in close parallelism to the requirements. Such requirements engineering also results in direct and automatic translation to implementation, minimising software faults. We present our method and illustrated with a classical example. We also compare our approach with other state diagram methods, as well as Petri nets and Behavior Trees.

Plausible Logic Facilitates Engineering the Behaviour of Autonomous Robots

In this paper we extend finite state machines to allow expressions in Plausible Logic for labelling transitions. As a result, we enable the design of behaviours that incorporate non-monotonic reasoning with a high-level software development tool. Using a cognitive software architecture that supports the efficient implementation of a developing/ programming environment, we automatically translate graphical designs of behaviour into executables that run on board autonomous robots. The graphical designs are obtained by demonstrating the transformation of the state machine into a Behavior Tree does not lose information and enhances modularisation of logic descriptions. We illustrate this with a description of the rapid development of the behaviour of a friendly poker player on an Aibo that interacts with humans.