Reearch into virtual environments on the one hand and artificial intelligence and artificial life on the other has largely been carried out by two different groups of people with different preoccupations and interests, but some convergence is now apparent between the two fields. Applications in which activity independent of the user takes place --- involving crowds or other agents --- are beginning to be tackled, while synthetic agents, virtual humans and computer pets are all areas in which techniqes from the two fields require strong integration. The two communities have much to learn from each other if wheels are not to be reinvented on both sides. This paper reviews the issues arising from combining artificial intelligence and artificial life techniques with those of virtual environments to produce just such intelligent virtual environments. The discussion is illustrated with examples that include environments providing knowledge to direct or assist the user rather than r...

There is a growing body of work that concentrates on theoretical aspects of agents and multi-agent systems, and a complementary body of work concerned with building practical systems. However, the two have typically been unrelated. This gap between the theory and practice of intelligent agents has only relatively recently begun to be addressed. In this paper we describe the construction of an agent simulation environment that is based strongly on a formal theory of agent systems, but which is intended to serve in exactly this way as a basis for practical development. The paper briefly introduces the theory, then describes the system and the simple reactive agents built with it, but most importantly shows how it reflects the theoretical framework and how it facilitates incremental agent design and implementation. Using this example as a case-study, some possibilities for a methodology for the development of agent systems are discussed.

Social behaviour arises as a result of individual agents cooperating with each other so as to exploit the resources available in a rich and dynamic multi-agent domain. If agents are to make use of others to help them in their tasks, such social behaviour is critical. Underlying this cooperation is the transfer or adoption of goals from one agent to another, a subtle and complex process that depends on the nature of the agents involved. In this paper we analyse this process by building upon a hierarchy previously constructed to define objects, agents and autonomous agents. We describe the motivated self-generation of goals that defines agent autonomy and the adoption of goals between agents that enables social behaviour. Then we consider three classes of goal adoption by objects, agents and autonomous agents. The first of these is merely a question of instantiation, the second requires an understanding of the relationship of the agent to others that are engaging it, and the third amounts to a question of negotiation or persuasion.

This thesis presents a study of the provision of emotions for artificial agents with the ultimate aim of enhancing their autonomy, i.e. making them more flexible, robust and self-sufficient. In recent years, the importance of emotions and their assistance to cognition has been increasingly acknowledged. Emotions are no longer considered undesirable or simply useless. Their role in various aspects of human and animal cognition like perception, attention, memory, decision-making and social interaction has been recognised as essential. The importance of emotions is much more evident insocial interaction and therefore much of the emotions research done in artificial systems focuses on the expression and recognition of emotions. However, recent neurophysiological research suggests that emotions also play a crucial part in cognition itself. This thesis investigates ways in which artificial emotions can improve autonomous behaviour in the domain of a simple, but complete, solitary learning agent. For this purpose, a non-symbolic emotion model was designed and implemented. It takes the form of a recurrent artificial neural network where emotions influence the perception

The fact that emotions are considered to be essential to human reasoning suggests that they might play an important role in autonomous robots. Experimental

In the paper we discuss variable and generative forms of autonomy. Variable autonomy is discussed in terms of the practicalities in designing autonomous agents, dealing as it does with the notion of degrees of autonomy and hence issues of agent control. The major part of the paper discusses an absolute, theoretically grounded notion of autonomy: the ability to generate one's own goals. This theoretical account of autonomy is embedded in the larger SMART framework and is intimately linked with the issue of motivation. Autonomous agents are motivated agents in that for the generation of goals an agent needs a set of higher order, non-derivative sources of action, or in our terminology, motivations. Autonomous agents in the SMART framework form the basis and source of action in multi-agent systems, which can thus propagate through the other entities in the system, such as non-autonomous agents and objects. We conclude with a discussion regarding the situations an autonomous agent would be willing to relinquish its autonomy thus linking the generative and variable notions of autonomy.

Autonomy is a very important property for a robot to have, yet implementing it in a robot is far from trivial, particularly when one requires the meaning of autonomy to include self-motivation, instead of mere automaticity. The fact that emotions are considered to be essential to human reasoning and human motivation in particular, suggests that they might play an important role in robot autonomy. The purpose of the work reported here is to know if and how emotions can help a robot in achieving autonomy. Experimental work was done in a simulated robot that adapts to its environment through the use of reinforcement learning. Results suggest that emotions can be useful in dividing the task in smaller manageable problems by focusing attention on the relevant features of the task at any one time.

Humour is a valid subject for research in artificial intelligence, as it is one of the more complex of human behaviours. Although philosophers and others have discussed humour for centuries, it is only very recently that computational work has begun in this field, so the state of the art is still rather basic. Much of the research has concentrated on humour expressed verbally, and there has been some emphasis on models based on "incongruity". Actual implementations have involved puns of very limited forms. It is not clear that computerised jokes could enhance user interfaces in the near future, but there is a role for computer modelling in testing symbolic accounts of the structure of humorous texts. A major problem is the need for a humour-processing program to have knowledge of the world, and reasoning abilities.

Contents Abstract xiii Declaration xiv Copyright xv Declaration of honesty xvi

This paper describes current work from an ongoing investigation into the computational modelling of emotion and motivation. The impetus for this research can be found in the work on the computational modelling of motivation. The impetus for that in turn is the control state approach to modelling cognition. Recent research has focussed on the use of cellular automata to provide a foundation for an emotion engine. This emotion engine is to be placed at the core of an agent architecture. Earlier work used simple tokens, for example fuzzy-valued symbols such as angry or very angry, in motivational structures to act as referents to specific emotional states. Having revisited concepts underlying the nature of agency, a number of simple multi-agent scenarios are being investigated as experimental frameworks for developing the theory, designs and implementations. 1