Lydie Edward

An Action Selection Architecture for Autonomous Virtual Agents

Many day-to-day applications involve autonomous agents. When designing autonomous agents, the problem of selecting actions must be considered, as it governs decision making at all times. In this paper we describe how we designed intelligent virtual cognitive agents representing operators performing tasks in a high-risk plant. They must respond to expected as well as unexpected events. The reasoning system controlling the agent behavior is designed to exhibit human behavior. Agents must be able to plan their actions according to their perception, their beliefs and their goals. We developed a planning system that interleaves plan construction and plan execution. The new planner, called AATP, produces a plan according to agent goal.

Cognitive Modeling of Virtual Autonomous Intelligent Agents Integrating Human Factors

We introduce a cognitive architecture combining human factors, human cognitive modeling, decision, perception and action selection based on hierarchical task networks. We use a cognitive model that comes from phycology science to define the control mode of a virtual agent. This control mode influences the agent decisions as well as its perception, its knowledge, or its goals. It depends on its physical and cognitive states and on its personality. We determine which factors influence the agent control mode and use them to generate various behaviors according to the situational constraints. We propose a new approach for modeling the environment and new algorithms to enable our virtual characters to evolve in the environment dynamically with a certain level of credibility.

An Ontology for Managing a Virtual Environment for Risk Prevention

In this paper we present how we combine artificial intelligence and knowledge engineering to provide an ontology inspired approach in order to manage a virtual environment for risk prevention. In the virtual environment, different entities cohabit: virtual operators represented by cognitive agents and the learner’s avatar that represents a real operator. They can interact with the objects through actions and modify their state. It is therefore useful to have on one hand a system that specifically defines the framework in which the interactions or actions are allowed and on the other hand a representation of the knowledge involved in such interactions. The ontology models the objects of the environment, the agents operating in the environment, the events that can happen, the actions and the relations between all such concepts. We proposed an agent, COLOMBO, containing the developed ontology and a set of reasoning rules.

Une architecture cognitive pour la simulation de comportements d'agents autonomes: Intégration de facteurs physiques, physiologiques et de personnalité

Our goal is to model the behavior of autonomous agents evolving in a disrupted virtual environment. To improve the credibility of their behaviors we propose to model their cognitive states by relying on models developed by cognitive scientists. Autonomous agents have deviated behaviors according to their state. We introduce a cognitive agent architecture based on the BDI model enhanced with a model integrating personality, physical and physiological characteristics, which influences agent decisions and action selections.

Virtual Autonomous Agents in an Informed Environment for Risk Prevention

This paper presents the work to enhance virtual agents with the abilities of interacting with the environment and planning based on goals and internal states. We describe how we designed intelligent virtual agents that are able to perform tasks on a high-risk plant and that respond to expected as well as unexpected events. We proposed an informed virtual environment where agents are able to interact with other agents or avatars. By enhancing their reasoning abilities with cognitive rules and the environment with knowledge the virtual agents are able to exhibit deviated behaviors.

Human factors affecting decision in virtual operator reasoning

In this paper we present our research on virtual operator cognitive modeling and reasoning. They operate in a virtual environment for risk prevention. We want to model the influence of their cognitive states and personality on their decisions and actions in the environment. The purpose of our research is to design a system that generates behaviorbased errors to support learning and risk prevention. It uses new mechanisms taking into account human factors and human behavior model regarding risky situations. We make use of the COCOM and CREAM model describing the different states or control mode in which an agent can be regarding to a lack of time or a temporal pressure. A challenge is to implement these models to produce the expected flexible, contextual and erroneous behaviors in both normal and constrained working conditions.