Throughout history education has been restricted to a relatively small percentage of the world's population. The cause can be attributed to a number of factors; how- ever, it has been chiefly due to excessive cost. As we enter the information age it becomes conceivable to make education freely available to anyone, anywhere, any- time. The Intelligent Tutoring System is an automated teaching system designed to improve through experience, eventually learning to tailor its teaching to perfectly match each individual student's needs and preferences. In this dissertation we describe a template which we use for building problem-oriented skill teaching intelligent tutoring systems based on a Dynamic Bayes network framework. We present two case studies in which the template is adapted to very different teaching domains, documenting in each case the process of building, training, and testing the resulting ITS. In both case studies, the performance of the ITS is validated through human subject experiments. The results of these studies show that our template is a viable technique for designing ITSs that teach in skill based domains. We also show that, while conducting artificial intelligence research on the design of an ITS and collecting data for use in that regard, we can concurrently run educational research experiments. We find that the two are quite inextricably tied and that showing good general results regarding the performance of the ITS is not sufficient; a strong understanding of the experience of the students is also required. We report some interesting results covering the effect of choice in learning and a gender bias that shows up in our tutoring system.

No description available.

No description available.

No description available.

This thesis addresses the problem of intelligent control of autonomous mobile robots, particularly under circumstances unforeseen by the designer. As the range of applications for autonomous robots widens and increasingly includes operation in unknown environments (exploration) and tasks which are not clearly specifiable a priori (maintenance work), this question is becoming more and more important. It is argued that in order to achieve such flexibility in unforeseen situations it is necessary to equip a mobile robot with the ability to autonomously acquire the necessary task achieving competences, through interaction with the world. Using mobile robots equipped with self-organising, behaviour-based controllers,experiments in the autonomous acquisition of motor competences and navigational skills were conducted to investigate the viability of this approach. A controller architecture is presented that allows extremely fast acquisition of motor competence such as obstacle avoidance, wall and corridor following and deadend escape: these skills are obtained in less than five learning steps,performed in under one minute of real time. This is considerably faster than previous approaches. Because the effective wiring between sensors and actuators is determined autonomously by the robot, sensors and actuators may initially be wired up arbitrarily,which reduces the risk of human error during the setting up phase of the robot. For the first time it was demonstrated that robots also become able to autonomously recover from unforeseen situations such as changes in the robot's morphology, the environment or the task. Rule-based approaches to error recovery obviously cannot offer recovery from unforeseen errors,as error situations covered by such approaches have to be identified beforehand. A robust and fast map building architecture is presented that enables mobile robots to autonomously construct internal representations of their environment, using self-organising feature maps. After a short training time the robots are able to use these self-organising feature maps successfully for location recognition. For the first time the staged acquisition of multiple competences in mobile robots is presented. First obtaining fundamental motor competences such as wall following and deadend escape (primary skills), the robots use these in a second stage to learn higher levels of competence such as the navigational task of location recognition (secondary skills). Besides laying the foundation of autonomous, staged acquisition of high level competences, this approach has the interesting property of securely grounding secondary skills in the robot's own experience, as these secondary skills are defined in terms of the primary ones.

Street lighting is an important aspect of infrastructure in terms of both safety and comfort, but it also consumes a lot of energy. Unused light is a waste of energy, and without any form of control of the street lighting, this problem will continue to increase along with the expansion of road networks. The aim of this thesis is to propose an intelligent control system for street lighting that can adapt to the velocity of individual road users, to investigate if this could provide ways to improve the efficiency of street lighting. Previous control approaches include systems based on ambient light intensity or presence of road users, but no studies were found in which illumination adapts to the velocity of road users. The project involves three main steps, including a literature review, a system implementation and evaluation. In the proposed system, street lights cooperate to detect road users and calculate their velocities in order to adapt the illumination and make it follow their movement. It can be concluded from the evaluation results that the velocity readings help further optimize the illumination control in comparison to systems that do not consider velocity. The velocity readings make it possible to only illuminate the roadway in the direction of travel, while also adapting the distance of illumination to the recorded speed. The proposed control scheme is considered a viable solution for reducing the amount of unused light, consequently reducing the energy consumption of street lighting.

In this thesis I explore how older people make use of, and interact with, their physical environment in home and near-by settings to manage cognitive situations, specifically prospective memory situations. Older adults have in past research been shown to perform better on prospective memory in real-life settings than what findings in laboratory-like settings predict. An explanation for this paradox is that older adults has a more developed skill of using the environment for prospective memory than younger adults. However, research investigating this explanation has primarily been based on self-reports. I contribute to the understanding of this skill by doing two related things. First I introduce distributed cognition, a theoretical perspective that primarily has been used within professional and socio-technical environments, to the research field of prospective memory in everyday life. Second I present a cognitive ethnography conducted during two years across eight home, and near-by, environments and old-age retired persons, for which I have used theoretical concepts from distributed cognition to analyze observations. The analysis shows rich variations in how participants use common cultural cognitive tools, invent their own cognitive tools, deliberately and incidentally shape more or less functional spaces, make use of other physical features, orient themselves toward and make sense of cognitive resources. I complement both prospective memory and distributed cognition research by describing both the intelligent shaping and use of space. Furthermore, by taking a distributed cognitive perspective I show that prospective memory processes in home environments involve properties, and the management, of a multipurpose environment. Altogether this supports the understanding of distributed cognition as a perspective on all cognition. Distributed cognition is not a reflection of particular work practices, instead it is a formulation of the general features of human cognition. Prospective memory in everyday life can be understood as an ability persons have. However, in this thesis I show that prospective memory can also be understood as a process that takes place between persons, arrangements of space, and tools.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

M.Sc. / The use of computer systems has changed over the years. Modern computer systems operate in an environment that is open, distributed and heterogeneous. They have the capability of locating information stored in remote locations and satisfying the interests and objectives of different users. However, the increase in user demands and the complexity of computers and information systems has caused research to focus on multi-agent systems as a solution to address these demands and complexities. The dissertation deals with the study of single agents and multi-agent systems. The study focuses on the concepts of agents, agent architecture and multi-agent systems. In addition to the study, a taxonomy for specialised agents is proposed. The taxonomy aims at classifying agent-based systems applied in the industry for addressing specific problems. In order to achieve this, a broad survey on agent-based systems in the industry was conducted. The areas under considerations were the financial, health, agricultural, aviation and the information technology sectors. The following dimensions were used to identify the agents in the specific area: • Which application domain is the multi-agent system designed for, developed and deployed in? • What is the specific task or problem the agents are designed to solve? • Do the agents have core or advanced agent attributes in general? The taxonomy is important because agent-based systems are becoming common in the industry and are suitable to address issues (such as locating distributed information and addressing specific needs of computer system users) of open, distributed and heterogeneous computer environments.

M.Sc. / This dissertation investigates how auctioneer agents can maximise the revenue of an auction. Auctions are an effective solution to agent negotiation because of their simplicity. They are therefore the most widely used approach to agent negotiation. A review of auction theory proves that auction revenue is influenced by factors such as the auction format and the auction parameters. The optimal auction format and parameters are dependent on the bidders and the auction environment. A study of internet auctions confirms the findings of auction theory regarding the importance of the format as well as the parameters in an auction. There has been significant research on the strategies bidding agents can adopt in order to maximise their utility. There is little research, however, on the strategies auctioneer agents can adopt to maximise auction revenue. In this study, a survey is done of the related auctioneer agents in research that maximise auction revenue. The shortcomings of these are identified. The aim of this research is to develop an adaptive auctioneer agent that can maximise the revenue of an auction. The proposed solution is the revenue maximising adaptive auctioneer (RMAA) agent. The RMAA agent consists of three components: the auction manager, the auction format selector (AFS) and the auction parameter selector (APS). The AFS employs a customised zeroth-level classifier system to select the auction format based on the auction environment. The APS employs reinforcement learning combined with a local hill climbing algorithm to learn what the most profitable values are for auction parameters, such as the reserve price and the bid increment. The auction manager automates the auction and coordinates the AFS and APS. It implements the four traditional auction formats, namely the English auction, the Dutch auction, the Vickrey auction and the first-price sealed-bid auction. Experiments conducted to evaluate the effectiveness of the RMAA agent prove that this agent can consistently generate more revenue than a non-adaptive auctioneer agent. The contribution of this research is an auctioneer agent that adapts both the auction format and the auction parameters according to the environment in order to maximise the auction revenue.