Una propuesta de modelado del estudiante basada en ontologías y diagnóstico pedagógico-cognitivo no monótono

Los recientes avances tecnológicos han encontrado un potencial campo de explotación en la educación asistida por computador. A finales de los años 90 surgió un nuevo campo de investigación denominado Entornos Virtuales Inteligentes para el Entrenamiento y/o Enseñanza (EVIEs), que combinan dos áreas de gran complejidad: Los Entornos Virtuales (EVs) y los Sistemas de Tutoría Inteligente (STIs). De este modo, los beneficios de los entornos 3D (simulación de entornos de alto riesgo o entornos de difícil uso, etc.) pueden combinarse con aquéllos de un STIs (personalización de materias y presentaciones, adaptación de la estrategia de tutoría a las necesidades del estudiante, etc.) para proporcionar soluciones educativas/de entrenamiento con valores añadidos. El Modelo del Estudiante, núcleo de un SIT, representa el conocimiento y características del estudiante, y refleja el proceso de razonamiento del estudiante. Su complejidad es incluso superior cuando los STIs se aplican a EVs porque las nuevas posibilidades de interacción proporcionadas por estos entornos deben considerarse como nuevos elementos de información clave para el modelado del estudiante, incidiendo en todo el proceso educativo: el camino seguido por el estudiante durante su navegación a través de escenarios 3D; el comportamiento no verbal tal como la dirección de la mirada; nuevos tipos de pistas e instrucciones que el módulo de tutoría puede proporcionar al estudiante; nuevos tipos de preguntas que el estudiante puede formular, etc. Por consiguiente, es necesario que la estructura de los STIs, embebida en el EVIE, se enriquezca con estos aspectos, mientras mantiene una estructura clara, estructurada, y bien definida. La mayoría de las aproximaciones al Modelo del Estudiante en STIs y en IVETs no consideran una taxonomía de posibles conocimientos acerca del estudiante suficientemente completa. Además, la mayoría de ellas sólo tienen validez en ciertos dominios o es difícil su adaptación a diferentes STIs. Para vencer estas limitaciones, hemos propuesto, en el marco de esta tesis doctoral, un nuevo mecanismo de Modelado del Estudiante basado en la Ingeniería Ontológica e inspirado en principios pedagógicos, con un modelo de datos sobre el estudiante amplio y flexible que facilita su adaptación y extensión para diferentes STIs y aplicaciones de aprendizaje, además de un método de diagnóstico con capacidades de razonamiento no monótono. El método de diagnóstico es capaz de inferir el estado de los objetivos de aprendizaje contenidos en el SIT y, a partir de él, el estado de los conocimientos del estudiante durante su proceso de aprendizaje. La aproximación almodelado del estudiante propuesta ha sido implementada e integrada en un agente software (el agente de modelado del estudiante) dentro de una plataforma software existente para el desarrollo de EVIEs denominadaMAEVIF. Esta plataforma ha sido diseñada para ser fácilmente configurable para diferentes aplicaciones de aprendizaje. El modelado del estudiante presentado ha sido implementado e instanciado para dos tipos de entornos de aprendizaje: uno para aprendizaje del uso de interfaces gráficas de usuario en una aplicación software y para un Entorno Virtual para entrenamiento procedimental. Además, se ha desarrollado una metodología para guiar en la aplicación del esta aproximación de modelado del estudiante a cada sistema concreto.---ABSTRACT---Recent technological advances have found a potential field of exploitation in computeraided education. At the end of the 90’s a new research field emerged, the so-called Intelligent Virtual Environments for Training and/or Education (IVETs), which combines two areas of great complexity: Virtual Environments (VE) and Intelligent Tutoring Systems (ITS). In this way, the benefits of 3D environments (simulation of high risk or difficult-to-use environments, etc.) may be combined with those of an ITS (content and presentation customization, adaptation of the tutoring strategy to the student requirements, etc.) in order to provide added value educational/training solutions. The StudentModel, core of an ITS, represents the student’s knowledge and characteristics, and reflects the student’s reasoning process. Its complexity is even higher when the ITSs are applied on VEs because the new interaction possibilities offered by these environments must be considered as new key information pieces for student modelling, impacting all the educational process: the path followed by the student during their navigation through 3D scenarios; non-verbal behavior such as gaze direction; new types of hints or instructions that the tutoring module can provide to the student; new question types that the student can ask, etc. Thus, it is necessary for the ITS structure, which is embedded in the IVET, to be enriched by these aspects, while keeping a clear, structured and well defined architecture. Most approaches to SM on ITSs and IVETs don’t consider a complete enough taxonomy of possible knowledge about the student. In addition, most of them have validity only in certain domains or they are hard to be adapted for different ITSs. In order to overcome these limitations, we have proposed, in the framework of this doctoral research project, a newStudentModeling mechanism that is based onOntological Engineering and inspired on pedagogical principles, with a wide and flexible data model about the student that facilitates its adaptation and extension to different ITSs and learning applications, as well as a rich diagnosis method with non-monotonic reasoning capacities. The diagnosis method is able to infer the state of the learning objectives encompassed by the ITS and, fromit, the student’s knowledge state during the student’s process of learning. The proposed student modelling approach has been implemented and integrated in a software agent (the student modeling agent) within an existing software platform for the development of IVETs called MAEVIF. This platform was designed to be easily configurable for different learning applications. The proposed student modeling has been implemented and it has been instantiated for two types of learning environments: one for learning to use the graphical user interface of a software application and a Virtual Environment for procedural training. In addition, a methodology to guide on the application of this student modeling approach to each specific system has been developed.

Especificación del comportamiento de un avatar en un mundo virtual mediante una ontología

El presente trabajo describe la construcción de una aplicación que controla a un Non Player Character (NPC), en un mundo virtual. La aplicación desarrollada, que tiene como nombre BotManager, realiza dos tareas fundamentales: 1) conectarse al repositorio de conocimiento, que en esta implementación es una ontología expresada en OWL, para obtener las acciones que debe realizar el NPC dentro del mundo virtual; y 2) ordenar al NPC que realice estas acciones en un mundo virtual creado con la plataforma OpenSimulator. BotManager puede tener variadas aplicaciones, por lo tanto puede ser usada como complemento en mundos virtuales aplicados a la educación, simulación, ocio, etc. Ahora bien, la principal razón que motivó el desarrollo del BotManager fue la de crear un sistema de demostración automática de tareas en un mundo virtual destinado a la educación/ entrenamiento. De esta forma, un Sistema Inteligente de Tutoría integrado con un mundo virtual podría demostrar paso a paso a un estudiante cómo realizar una tarea en el mundo virtual. La ontología que lee el BotManager extiende la ontología propuesta en la tesis “Una propuesta de modelado del estudiante basada en ontologías y diagnóstico pedagógico-cognitivo no monótono” de Julia Parraga en el 2011 (Ontología de Julia). La construcción y las pruebas del BotManager se llevaron a cabo en tres etapas: 1) creación de la Ontología de Acciones del NPC que extiende la Ontología de Julia; 2) diseño e implementación de la aplicación en C# que lee la ontología que contiene el plan de acción del NPC, y ordena al NPC realizar las acciones en el mundo virtual; y 3) pruebas de la aplicación con la práctica “preparación de una taza de cafe”, que es parte de un Laboratorio Virtual de Biotecnología. El BotManager se ha diseñado como una aplicación cliente que se conecta a un servidor de Open- Simulator. Por lo tanto, puede ejecutarse en una máquina distinta a la del servidor. Asimismo, en la implementación del BotManager se ha utilizado una librería gratuita denominada LibOpenMetaverse que permite controlar un NPC de forma remota.---ABSTRACT---This paper describes the construction of an application that controls a Non Player Character (NPC), in a virtual world. The application developed, called BotManager, performs two main tasks: 1) the connection to the repository of knowledge, which in this implementation is an ontology expressed in OWL, and retrieving the actions to be performed by the NPC within the virtual world; and 2) commanding the NPC to perform these actions in a virtual world created with the OpenSimulator platform. BotManager can have diverse applications, therefore it can be used as a complement in virtual worlds applied to education, simulation, entertainment, etc. However, the main reason behind the development of BotManager was to create an automatic demonstration of tasks in a virtual world for education / training. Thus, a virtual world integrated with an Intelligent Tutoring Systems could demonstrate step by step to a student how to perform a task in the virtual world. The ontology used by the BotManager extends ontology proposed in the thesis “A proposal for modeling ontologies based student and not monotonous teaching-cognitive diagnosis” by Julia Parraga in 2011 (Julia’s Ontology). Construction and testing of BotManager were conducted in three stages: 1) creation of the NPC Actions Ontology by extending the Julia’s Ontology; 2) design and implementation of the application in C# that reads the ontology containing the plan of action of the NPC, and commands the NPC to perform the read plan in the virtual world; and 3) testing of the application with the practice “preparing a cup of coffee”, which is part of a Virtual Laboratory of Biotechnology. The BotManager has been designed as a client application that connects to an OpenSimulator server. Therefore, it can run on a different machine to the server. To implement the BotManager we have used a free library called libopenmetaverse that allows us to control a NPC remotely.

Cross-entropy optimization for scaling factors of a fuzzy controller: a see-and-avoid approach for unmanned aerial systems

The Cross-Entropy (CE) is an efficient method for the estimation of rare-event probabilities and combinatorial optimization. This work presents a novel approach of the CE for optimization of a Soft-Computing controller. A Fuzzy controller was designed to command an unmanned aerial system (UAS) for avoiding collision task. The only sensor used to accomplish this task was a forward camera. The CE is used to reach a near-optimal controller by modifying the scaling factors of the controller inputs. The optimization was realized using the ROS-Gazebo simulation system. In order to evaluate the optimization a big amount of tests were carried out with a real quadcopter.

Homography-based ground plane detection using a single on-board camera

This study presents a robust method for ground plane detection in vision-based systems with a non-stationary camera. The proposed method is based on the reliable estimation of the homography between ground planes in successive images. This homography is computed using a feature matching approach, which in contrast to classical approaches to on-board motion estimation does not require explicit ego-motion calculation. As opposed to it, a novel homography calculation method based on a linear estimation framework is presented. This framework provides predictions of the ground plane transformation matrix that are dynamically updated with new measurements. The method is specially suited for challenging environments, in particular traffic scenarios, in which the information is scarce and the homography computed from the images is usually inaccurate or erroneous. The proposed estimation framework is able to remove erroneous measurements and to correct those that are inaccurate, hence producing a reliable homography estimate at each instant. It is based on the evaluation of the difference between the predicted and the observed transformations, measured according to the spectral norm of the associated matrix of differences. Moreover, an example is provided on how to use the information extracted from ground plane estimation to achieve object detection and tracking. The method has been successfully demonstrated for the detection of moving vehicles in traffic environments.

Detection and Tracking of Traffic Signs Using a Recursive Bayesian Decision Framework

In this paper we propose a new method for the automatic detection and tracking of road traffic signs using an on-board single camera. This method aims to increase the reliability of the detections such that it can boost the performance of any traffic sign recognition scheme. The proposed approach exploits a combination of different features, such as color, appearance, and tracking information. This information is introduced into a recursive Bayesian decision framework, in which prior probabilities are dynamically adapted to tracking results. This decision scheme obtains a number of candidate regions in the image, according to their HS (Hue-Saturation). Finally, a Kalman filter with an adaptive noise tuning provides the required time and spatial coherence to the estimates. Results have shown that the proposed method achieves high detection rates in challenging scenarios, including illumination changes, rapid motion and significant perspective distortion

Two Novel Methods For The Determination Of The Number Of Components In Independent Components Analysis Models

Independent Components Analysis is a Blind Source Separation method that aims to find the pure source signals mixed together in unknown proportions in the observed signals under study. It does this by searching for factors which are mutually statistically independent. It can thus be classified among the latent-variable based methods. Like other methods based on latent variables, a careful investigation has to be carried out to find out which factors are significant and which are not. Therefore, it is important to dispose of a validation procedure to decide on the optimal number of independent components to include in the final model. This can be made complicated by the fact that two consecutive models may differ in the order and signs of similarly-indexed ICs. As well, the structure of the extracted sources can change as a function of the number of factors calculated. Two methods for determining the optimal number of ICs are proposed in this article and applied to simulated and real datasets to demonstrate their performance.

On-board and Ground Visual Pose Estimation Techniques for UAV Control

In this paper, two techniques to control UAVs (Unmanned Aerial Vehicles), based on visual information are presented. The first one is based on the detection and tracking of planar structures from an on-board camera, while the second one is based on the detection and 3D reconstruction of the position of the UAV based on an external camera system. Both strategies are tested with a VTOL (Vertical take-off and landing) UAV, and results show good behavior of the visual systems (precision in the estimation and frame rate) when estimating the helicopter¿s position and using the extracted information to control the UAV.

Automatic lateral control for unmanned vehicles via genetic algorithms

It is known that the techniques under the topic of Soft Computing have a strong capability of learning and cognition, as well as a good tolerance to uncertainty and imprecision. Due to these properties they can be applied successfully to Intelligent Vehicle Systems; ITS is a broad range of technologies and techniques that hold answers to many transportation problems. The unmannedcontrol of the steering wheel of a vehicle is one of the most important challenges facing researchers in this area. This paper presents a method to adjust automatically a fuzzy controller to manage the steering wheel of a mass-produced vehicle; to reach it, information about the car state while a human driver is handling the car is taken and used to adjust, via iterative geneticalgorithms an appropriated fuzzy controller. To evaluate the obtained controllers, it will be considered the performance obtained in the track following task, as well as the smoothness of the driving carried out.

Making Transport Safer: V2V-Based Automated Emergency Braking System

An important goal in the field of intelligent transportation systems (ITS) is to provide driving aids aimed at preventing accidents and reducing the number of traffic victims. The commonest traffic accidents in urban areas are due to sudden braking that demands a very fast response on the part of drivers. Attempts to solve this problem have motivated many ITS advances including the detection of the intention of surrounding cars using lasers, radars or cameras. However, this might not be enough to increase safety when there is a danger of collision. Vehicle to vehicle communications are needed to ensure that the other intentions of cars are also available. The article describes the development of a controller to perform an emergency stop via an electro-hydraulic braking system employed on dry asphalt. An original V2V communication scheme based on WiFi cards has been used for broadcasting positioning information to other vehicles. The reliability of the scheme has been theoretically analyzed to estimate its performance when the number of vehicles involved is much higher. This controller has been incorporated into the AUTOPIA program control for automatic cars. The system has been implemented in Citroën C3 Pluriel, and various tests were performed to evaluate its operation.

Energy- and environmentally efficient road management: The case of the Spanish motorway network

Transport is responsible for 41% of CO2 emissions in Spain, and around 65% of that figure is due to road traffic. Tolled motorways are currently managed according to economic criteria: minimizing operational costs and maximizing revenues from tolls. Within this framework, this paper develops a new methodology for managing motorways based on a target of maximum energy efficiency. It includes technological and demand-driven policies, which are applied to two case studies. Various conclusions emerge from this study. One is, that the use of intelligent payment systems is recommended; and another, is that the most sustainable policy would involve defining the most efficient strategy for each motorway section, including the maximum use of its capacity, the toll level which attracts the most vehicles, and the optimum speed limit for each type of vehicle.

Floating car data augmentation based on infrastructure sensors and neural networks

The development of new-generation intelligent vehicle technologies will lead to a better level of road safety and CO2 emission reductions. However, the weak point of all these systems is their need for comprehensive and reliable data. For traffic data acquisition, two sources are currently available: 1) infrastructure sensors and 2) floating vehicles. The former consists of a set of fixed point detectors installed in the roads, and the latter consists of the use of mobile probe vehicles as mobile sensors. However, both systems still have some deficiencies. The infrastructure sensors retrieve information fromstatic points of the road, which are spaced, in some cases, kilometers apart. This means that the picture of the actual traffic situation is not a real one. This deficiency is corrected by floating cars, which retrieve dynamic information on the traffic situation. Unfortunately, the number of floating data vehicles currently available is too small and insufficient to give a complete picture of the road traffic. In this paper, we present a floating car data (FCD) augmentation system that combines information fromfloating data vehicles and infrastructure sensors, and that, by using neural networks, is capable of incrementing the amount of FCD with virtual information. This system has been implemented and tested on actual roads, and the results show little difference between the data supplied by the floating vehicles and the virtual vehicles.

Rapid prototyping framework for visual control of autonomous micro aerial vehicles

Rapid prototyping environments can speed up the research of visual control algorithms. We have designed and implemented a software framework for fast prototyping of visual control algorithms for Micro Aerial Vehicles (MAV). We have applied a combination of a proxy-based network communication architecture and a custom Application Programming Interface. This allows multiple experimental configurations, like drone swarms or distributed processing of a drone’s video stream. Currently, the framework supports a low-cost MAV: the Parrot AR.Drone. Real tests have been performed on this platform and the results show comparatively low figures of the extra communication delay introduced by the framework, while adding new functionalities and flexibility to the selected drone. This implementation is open-source and can be downloaded from www.vision4uav.com/?q=VC4MAV-FW

Decarbonization of toll plazas: impact assessment of toll collection system management

Transport climate change impacts have become a worldwide concern. The use of Intelligent Transport Systems (ITS) could contribute to a more effective use of resources in toll road networks. Management of toll plazas is central to the reduction of greenhouse gas (GHG) emissions, as it is there that bottlenecks and congestion occur. This study focuses on management strategies aimed at reducing climate change impacts of toll plazas by managing toll collection systems. These strategies are based on the use of different collection system technologies – Electronic Toll Collection (ETC) and Open Road Tolling (ORT) – and on queue management. The carbon footprint of various toll plazas is determined by a proposed integrated methodology which estimates the carbon dioxide (CO2) emissions of the different operational stages at toll plazas (deceleration, service time, acceleration, and queuing) for the different toll collection systems. To validate the methodology, two main-line toll plazas of a Spanish toll highway were evaluated. The findings reveal that the application of new technologies to toll collection systems is an effective management strategy from an environmental point of view. The case studies revealed that ORT systems lead to savings of up to 70% of CO2 emissions at toll plazas, while ETC systems save 20% comparing to the manual ones. Furthermore, queue management can offer a 16% emissions savings when queue time is reduced by 116 seconds. The integrated methodology provides an efficient environmental management tool for toll plazas. The use of new technologies is the future of the decarbonization of toll plazas.

Genetic fuzzy-based steering wheel controller using a mass-produced car

Intelligent Transportation Systems (ITS) cover a broad range of methods and technologies that provide answers to many problems of transportation. Unmanned control of the steering wheel is one of the most important challenges facing researchers in this area. This paper presents a method to adjust automatically a fuzzy controller to manage the steering wheel of a mass-produced vehicle to reproduce the steering of a human driver. To this end, information is recorded about the car's state while being driven by human drivers and used to obtain, via genetic algorithms, appropriate fuzzy controllers that can drive the car in the way that humans do. These controllers have satisfy two main objectives: to reproduce the human behavior, and to provide smooth actions to ensure comfortable driving. Finally, the results of automated driving on a test circuit are presented, showing both good route tracking (similar to the performance obtained by persons in the same task) and smooth driving.

Arquitectura de Control para la Conducción Autónoma de Vehículos en Entornos Urbanos y Autovías

Hoy en día, el desarrollo tecnológico en el campo de los sistemas inteligentes de transporte (ITS por sus siglas en inglés) ha permitido dotar a los vehículos con diversos sistemas de ayuda a la conducción (ADAS, del inglés advanced driver assistance system), mejorando la experiencia y seguridad de los pasajeros, en especial del conductor. La mayor parte de estos sistemas están pensados para advertir al conductor sobre ciertas situaciones de riesgo, como la salida involuntaria del carril o la proximidad de obstáculos en el camino. No obstante, también podemos encontrar sistemas que van un paso más allá y son capaces de cooperar con el conductor en el control del vehículo o incluso relegarlos de algunas tareas tediosas. Es en este último grupo donde se encuentran los sistemas de control electrónico de estabilidad (ESP - Electronic Stability Program), el antibloqueo de frenos (ABS - Anti-lock Braking System), el control de crucero (CC - Cruise Control) y los más recientes sistemas de aparcamiento asistido. Continuando con esta línea de desarrollo, el paso siguiente consiste en la supresión del conductor humano, desarrollando sistemas que sean capaces de conducir un vehículo de forma autónoma y con un rendimiento superior al del conductor. En este trabajo se presenta, en primer lugar, una arquitectura de control para la automatización de vehículos. Esta se compone de distintos componentes de hardware y software, agrupados de acuerdo a su función principal. El diseño de la arquitectura parte del trabajo previo desarrollado por el Programa AUTOPIA, aunque introduce notables aportaciones en cuanto a la eficiencia, robustez y escalabilidad del sistema. Ahondando un poco más en detalle, debemos resaltar el desarrollo de un algoritmo de localización basado en enjambres de partículas. Este está planteado como un método de filtrado y fusión de la información obtenida a partir de los distintos sensores embarcados en el vehículo, entre los que encontramos un receptor GPS (Global Positioning System), unidades de medición inercial (IMU – Inertial Measurement Unit) e información tomada directamente de los sensores embarcados por el fabricante, como la velocidad de las ruedas y posición del volante. Gracias a este método se ha conseguido resolver el problema de la localización, indispensable para el desarrollo de sistemas de conducción autónoma. Continuando con el trabajo de investigación, se ha estudiado la viabilidad de la aplicación de técnicas de aprendizaje y adaptación al diseño de controladores para el vehículo. Como punto de partida se emplea el método de Q-learning para la generación de un controlador borroso lateral sin ningún tipo de conocimiento previo. Posteriormente se presenta un método de ajuste on-line para la adaptación del control longitudinal ante perturbaciones impredecibles del entorno, como lo son los cambios en la inclinación del camino, fricción de las ruedas o peso de los ocupantes. Para finalizar, se presentan los resultados obtenidos durante un experimento de conducción autónoma en carreteras reales, el cual se llevó a cabo en el mes de Junio de 2012 desde la población de San Lorenzo de El Escorial hasta las instalaciones del Centro de Automática y Robótica (CAR) en Arganda del Rey. El principal objetivo tras esta demostración fue validar el funcionamiento, robustez y capacidad de la arquitectura propuesta para afrontar el problema de la conducción autónoma, bajo condiciones mucho más reales a las que se pueden alcanzar en las instalaciones de prueba. ABSTRACT Nowadays, the technological advances in the Intelligent Transportation Systems (ITS) field have led the development of several driving assistance systems (ADAS). These solutions are designed to improve the experience and security of all the passengers, especially the driver. For most of these systems, the main goal is to warn drivers about unexpected circumstances leading to risk situations such as involuntary lane departure or proximity to other vehicles. However, other ADAS go a step further, being able to cooperate with the driver in the control of the vehicle, or even overriding it on some tasks. Examples of this kind of systems are the anti-lock braking system (ABS), cruise control (CC) and the recently commercialised assisted parking systems. Within this research line, the next step is the development of systems able to replace the human drivers, improving the control and therefore, the safety and reliability of the vehicles. First of all, this dissertation presents a control architecture design for autonomous driving. It is made up of several hardware and software components, grouped according to their main function. The design of this architecture is based on the previous works carried out by the AUTOPIA Program, although notable improvements have been made regarding the efficiency, robustness and scalability of the system. It is also remarkable the work made on the development of a location algorithm for vehicles. The proposal is based on the emulation of the behaviour of biological swarms and its performance is similar to the well-known particle filters. The developed method combines information obtained from different sensors, including GPS, inertial measurement unit (IMU), and data from the original vehicle’s sensors on-board. Through this filtering algorithm the localization problem is properly managed, which is critical for the development of autonomous driving systems. The work deals also with the fuzzy control tuning system, a very time consuming task when done manually. An analysis of learning and adaptation techniques for the development of different controllers has been made. First, the Q-learning –a reinforcement learning method– has been applied to the generation of a lateral fuzzy controller from scratch. Subsequently, the development of an adaptation method for longitudinal control is presented. With this proposal, a final cruise control controller is able to deal with unpredictable environment disturbances, such as road slope, wheel’s friction or even occupants’ weight. As a testbed for the system, an autonomous driving experiment on real roads is presented. This experiment was carried out on June 2012, driving from San Lorenzo de El Escorial up to the Center for Automation and Robotics (CAR) facilities in Arganda del Rey. The main goal of the demonstration was validating the performance, robustness and viability of the proposed architecture to deal with the problem of autonomous driving under more demanding conditions than those achieved on closed test tracks.