Cette thèse en Cognitique se focalise sur la « Conception Centrée sur l'Humain » (Human Centred Design) de futures assistances à la conduite automobile, adaptées aux conducteurs âgés (ou Elderly Adapted Driver Assistance Systems).Pour ce faire, la démarche proposée repose sur une approche et une méthodologie pluridisciplinaire. Sur le plan ergonomique, il s'agit de mieux connaître les spécificitésde la population des conducteurs âgés, dans le but d'identifier des difficultés et des besoins en assistance. A cette fin, 76 conducteurs âgés (de 70 à 87 ans) ont conduitun véhicule instrumenté, immergé dans le trafic. Le corpus de données comporte2100 kilomètres de conduite et 1400 situations de conduite autoévaluées par lesconducteurs, complétés par 6 Focus Group (30 conducteurs âgés).Le second volet, relevant d'une démarche d'Ingénierie Cognitive, vise à concevoir et développer des fonctions de « monitorage » à partir du corpus de données. L'objectif est de disposer de modèles et de fonctions d'analyse temps-réel capables (1) de superviser l'activité de conduite des conducteurs âgés (2) en regard du contexte ou des risques situationnels, afin de (3) diagnostiquer des difficultés ou erreurs de conduite, à des fins d’adaptativité des assistances. Des fonctions de monitorage en lien avec les contrôles de base du véhicule (gestion de la vitesse, positionnement dans la voie et la gestion de l'espace inter-véhiculaire avant) sont développées. Sur cette base, des fonctions de monitorage plus intégrées pour l'aide aux franchissements d'intersections (Tourne-à-Gauche) et l'assistance à l'insertion sur voies rapides (et au changement de voie) sont également proposées. / This thesis in Cognitics presents a Human Centered Design approach for thedevelopment of future driving assistance systems dedicated to elderly drivers orElderly Adapted Driver Assistance Systems (E-ADAS).To do so, this work relies on a multi-disciplinary approach for data collection andanalysis. Regarding Ergonomics, the aim is to better understand the specificrequirements of this population in order to identify their actual difficulties and actualneeds of assistance. In this frame, 76 drivers (aged from 70 to 87 years old) took partto an on-the-road experiment, driving an instrumented car. The dataset includes2100 km of ecological driving data and 1400 auto-evaluated driving situations,completed by 6 Focus Groups (involving 30 elderly drivers).The second part of this research, relying on Cognitive Engineering, explores thedesign and implementation of monitoring functions based on the aforementioneddataset. The objective is to have real-time models and analytical functions, able to:(1) supervise the driving activity as realized by an elderly driver, (2) taking in toconsideration the driving context or situational risks (3) in order to detect difficulties ordriving errors. Beyond this thesis, these diagnostics will have to be integrated inassistive systems to better adapt their support to the specific needs of elderly drivers.Specific monitoring functions related to basic vehicle control (speed management,lane positioning and headway regulation) are presented. Based on these results,integrated monitoring functions for intersection crossings in Left-Turn manoeuver,highway merging assistance, and, more broadly, lane change assistance areintroduced.

Submitted by Juliana Cristina de Angelo (jul.angel@hotmail.com) on 2017-05-22T13:46:30Z No. of bitstreams: 1 Juliana Cristina de Angelo Repositorio.pdf: 20659453 bytes, checksum: 694ead497ef1e13185fc9690455644b5 (MD5) / Approved for entry into archive by LUIZA DE MENEZES ROMANETTO (luizamenezes@reitoria.unesp.br) on 2017-05-22T19:19:21Z (GMT) No. of bitstreams: 1 angelo_jc_me_bauru.pdf: 20659453 bytes, checksum: 694ead497ef1e13185fc9690455644b5 (MD5) / Made available in DSpace on 2017-05-22T19:19:21Z (GMT). No. of bitstreams: 1 angelo_jc_me_bauru.pdf: 20659453 bytes, checksum: 694ead497ef1e13185fc9690455644b5 (MD5) Previous issue date: 2017-03-23 / O objetivo deste estudo foi avaliar os efeitos da aptidão física e da velocidade do veículo sobre a aquisição da informação visual de motoristas experientes durante a condução simulada de automóveis. Quinze participantes fisicamente ativos, com idade de 37,46 ± 4,34 anos, IMC de 22,7 ± 2,57 kg/m2 e experiência de condução de 17,93 ± 4,06 anos, e quinze participantes sedentários, com idade de 30,66 ± 6,90 anos, IMC de 22,8 ± 3,87 kg/m2 e experiência de condução de 10,20 ± 5,08 anos, foram submetidos a uma tarefa de condução simulada de automóvel, de duração de três minutos, nas condições de velocidade 50-60, 80-90 e 110-120 Km/h, enquanto tiveram seus movimentos dos olhos e da cabeça e sua frequência cardíaca gravados. As variáveis dependentes adotadas foram número de fixações, duração média das fixações e sua variabilidade, tempo relativo de fixação, variâncias das posições horizontal e vertical do olhar, variâncias das posições e orientações tridimensionais da cabeça. Estes dados foram submetidos a uma análise de variância de Grupo (ativo, sedentário) por Velocidade (50-60, 80-90, 110-120 Km/h) com medidas repetidas no segundo fator. O questionário Baecke apresentou no score de exercícios físicos para classificação de aptidão física uma média de M = 3.53 (DP = 0.74) e o grupo sedentário M = 2.26 (DP = 0.67). A frequência cardíaca resultou em uma média significativamente afetada pela condição experimental, F(3,0, 83,2) = 4,51, p = 0,006 e pelo grupo, F(1, 28) = 5,50, p = 0,026. A variabilidade da duração das fixações dos movimentos dos olhos foi significativamente afetada pela condição velocidade, F(1,5 43,2) = 3,79, p =0,041 e a variabilidade da duração das fixações na velocidade 80-90 Km/h (M = 0,34, EP = 0, 17) foi significativamente maior (p = 0,019) do que na velocidade 110-120 Km/h (M = 0,31, EP = O, 17). Em síntese: o nível de aptidão física afetou significativamente a frequência cardíaca durante a condução simulada de automóvel em distintas velocidades; bem como a velocidade do veículo afetou significativamente a frequência cardíaca. Participantes ativos e sedentários foram semelhantes na aquisição de informação visual durante a condução simulada de automóvel em diferentes velocidades. / The purpose of this study was to evaluate the effects of physical fitness and driving velocity on the experienced drivers' visual information acquisition during simulated car driving. Fifteen physically fit participants, with age of 37,46 ± 4,34 years, BMI of 22,7 ± 2,57 kg/m2, and driving experience of 17,93 ± 4,06 years, and fifteen sedentary participants with age of 30,66 ± 6,90 years, BMI of 22,8 ± 3,87 kg/m2, and driving experience of 10,20 ± 5,08 years, were submitted to a simulated car driving task of 3-min duration, under the velocity conditions of 50-60, 80-90, and 110-120 Km/h, while had their gaze and head movements, and heart frequency recorded. The dependent variables adopted were number of fixations, mean fixation duration and its variability, relative fixation time, variances of horizontal and vertical gaze position, variances of tridimensional head position and orientation. These data were submitted to a Group (fit, sedentary) by driving velocity (50-60, 80-90, 110-120 Km/h) analysis of variance. The Baecke questionnaire presented a mean of M = 3.53 (SD = 0.74) and the sedentary group M = 2.26 (SD = 0.67) in the physical fitness score for physical fitness classification. The heart rate resulted in a mean significantly affected by the experimental condition, F (3.0, 83.2) = 4.51, p = 0.006 and by the group, F (1.28) = 5.50, p = 0.026. The variability of the fixation duration of the eye movements was significantly affected by the velocity condition, F (1.5 43.2) = 3.79, p = 0.041 and the variability of the fixation duration at the velocity 80-90 Km / h ( M = 0.34, SD = 0.17) was significantly higher (p = 0.019) than at the speed 110-120 km /h (M = 0.31, SD = 0.17). ln summary: the level of physical fitness significantly affected the heart rate during simulated car driving at different speeds; vehicle speed significantly affected heart rate. Active and sedentary participants were similar in acquiring visual information during simulated car driving at different speeds.

This master’s thesis is mainly focused on creating our own car driving simulator and basic data gathering from the driver’s input. To achieve this goal, the thesis includes introduction to man-machine systems and basic information about functions and runtime game engine employs. Research about commonly used open source game engines is also presented with a detailed focus on the engine we chose - Unreal Engine. In conclusion ofthis thesis, a full version of a car driving simulator is created with gathered data analysis.

Comprendre la genèse d’une situation de conduite requiert d’analyser les choixfaits par le conducteur au volant de son véhicule pendant l’activité de conduite, dans sacomplexité naturelle et dans sa dynamique située. Le LESCOT a développé le modèleCOSMODRIVE, fournissant un cadre conceptuel pour la simulation cognitive de l’activitéde conduite automobile. Pour exploiter ce modèle en simulation, il est nécessairede produire les connaissances liées à la situation de conduite sous forme d’un automatepar exemple. La conception d’un tel automate nécessite d’une part de disposer de donnéesissues de la conduite réelle, enregistrées sur un véhicule instrumenté et d’autrepart d’une expertise humaine pour les interpréter.Pour accompagner ce processus d’ingénierie des connaissances issues de l’analysed’activité, ce travail de thèse propose une méthode de découverte interactive deconnaissances à partir de traces d’activité. Les données de conduite automobile sontconsidérées comme des M-Traces, associant une sémantique explicite aux données,exploitées en tant que connaissances dans un Système à Base de Traces (SBT). Le SBTpermet de filtrer, transformer, reformuler et abstraire les séquences qui serviront à alimenterla synthèse de modèles automates de l’activité de conduite. Nous reprenons destechniques de fouille de workflow permettant de construire des automates (réseaux dePetri) à partir de logs. Ces techniques nécessitent des données complètes ou statistiquementreprésentatives. Or les données collectées à bord d’un véhicule en situationde conduite sont par nature des cas uniques, puisqu’aucune situation ne sera jamaisreproductible à l’identique, certaines situations particulièrement intéressantes pouvanten outre être très rarement observées. La gageure est alors de procéder à une forme degénéralisation sous la forme de modèle, à partir d’un nombre de cas limités, mais jugéspertinents, représentatifs, ou particulièrement révélateurs par des experts du domaine.Pour compléter la modélisation de telles situations, nous proposons donc de rendreinteractifs les algorithmes de synthèse de réseau de Petri à partir de traces, afin depermettre à des experts-analystes de guider ces algorithmes et de favoriser ainsi la découvertede connaissances pertinentes pour leur domaine d’expertise. Nous montreronscomment rendre interactifs l’algorithme α et l’algorithme α+ et comment généralisercette approche à d’autres algorithmes.Nous montrons comment l’utilisation d’un SBT et de la découverte interactived’automates impacte le cycle général de découverte de connaissances. Une méthodologieest proposée pour construire des modèles automates de l’activité de conduiteautomobile.Une étude de cas illustre la méthodologie en partant de données réelles de conduiteet en allant jusqu’à la construction de modèles avec un prototype logiciel développédans le cadre de cette thèse / Driving is a dynamic and complex activity. Understanding the origin of a driving situationrequires the analysis of the driver’s choices made while he/she drives. In addition,a driving situation has to be studied in its natural complexity and evolution. LESCOThas developed a model called COSMODRIVE, which provides a conceptual frameworkfor the cognitive simulation of the activity of car driving. In order to run themodel for a simulation, it is necessary to gather knowledge related to the driving situation,for example in the form of an automaton. The conception of such an automatonrequires : 1) the use of real data recorded in an instrumented car, and, 2) the use of humanexpertise to interpret these data. These data are considered in this thesis as activitytraces.The purpose of this thesis is to assist the Knowledge Engineering process of activityanalysis. The present thesis proposes a method to interactively discover knowledgefrom activity traces. For this purpose, data from car driving are considered as M-traces– which associate an explicit semantic to these data. This semantic is then used asknowledge in a Trace Based System. In a Trace Based System, M-traces can be filtered,transformed, reformulated, and abstracted. The resulting traces are then used as inputsin the production of an automaton model of the activity of driving. In this thesis,Workflow Mining techniques have been used to build automata (Petri nets) from logs.These techniques require complete or statistically representative data sets. However,data collected from instrumented vehicles are intrinsically unique, as no two drivingsituations will ever be identical. In addition, situations of particular interest, such ascritical situations, are rarely observed in instrumented vehicle studies. The challenge isthen to produce a model which is a form of generalisation from a limited set of cases,which have been judged by domain experts as being relevant and representative of whatactually happens.In the current thesis, algorithms synthesising Petri nets from traces have been madeinteractive, in order to achieve the modelling of such driving situations. This thenmakes it possible for experts to guide the algorithms and therefore to support the discoveryof knowledge relevant to the experts. The process involved in making the α-algorithm and the α+-algorithm interactive is discussed in the thesis in a way that canbe generalised to other algorithms.In addition, the current thesis illustrates how the use of a Trace Based System andthe interactive discovery of automata impacts the global cycle of Knowledge Discovery.A methodology is also proposed to build automaton models of the activity of cardriving. Finally, a case study is presented to illustrate how the proposed methodologycan be applied to real driving data in order to construct models with the softwaredeveloped in this thesis