Topological maps are light-weight, graphical representations of environments that are scalable and amenable to symbolic manipulation. Thus, they are well- suited for basic robot navigation applications, and also provide a representational basis for the procedural and semantic information needed for higher-level robotic tasks. However, their widespread use has been impeded in part by the lack of reliable, general purpose algorithms for their construction. In this dissertation, I present a probabilistic framework for the construction of topological maps that addresses topological ambiguity, is failure-aware, computa- tionally efficient, and can incorporate information from various sensing modalities. The framework addresses the two major problems of topological mapping, namely topological ambiguity and landmark detection. The underlying idea behind overcoming topological ambiguity is that the com- putation of the Bayesian posterior distribution over the space of topologies is an effective means of quantifying this ambiguity, caused due to perceptual aliasing and environment variability. Since the space of topologies is combinatorial, the posterior on it cannot be computed exactly. Instead, I introduce the concept of Probabilistic Topological Maps (PTMs), a sample-based representation that ap- proximates the posterior distribution over topologies given the available sensor measurements. Sampling algorithms for the efficient computation of PTMs are described. The PTM framework can be used with a wide variety of landmark detection schemes under mild assumptions. As part of the evaluation, I describe a novel landmark detection technique that makes use of the notion of "surprise" in mea- surements that the robot obtains, the underlying assumption being that landmarks are places in the environment that generate surprising measurements. The com- putation of surprise in a Bayesian framework is described and applied to various sensing modalities for the computation of PTMs. The PTM framework is the first instance of a probabilistic technique for topo- logical mapping that is systematic and comprehensive. It is especially relevant for future robotic applications which will need a sparse representation capable of accomodating higher level semantic knowledge. Results from experiments in real environments demonstrate that the framework can accomodate diverse sensors such as camera rigs and laser scanners in addition to odometry. Finally, results are pre- sented using various landmark detection schemes besides the surprise-based one.

Robotic haptics has been and continues to be an area of intense research, primarily in medical and exploration industries. This is due to an ability to provide high data throughput between human and machine. In medical applications, it is possible to detect and compensate errors such as a hand tremor in a surgeon. It is possible to apply scaling factors to assist in microsurgery situations, and can allow leading experts to perform procedures from anywhere on the globe. As part of a collaboration to develop a robotic method of femur fracture realignment between Auckland University, Auckland District Health Board, and Massey University, the project seeks to provide a haptic driven HMI for the realignment system. To reduce construction required, an existing manipulation arm (Mitsubishi RV-M1) is used as the hardware interface device. A new motor controller is designed to provide additional functionality as the standard controller provides no force control or real-time feedback of position. A software interface is developed (using version 3 of the C# programming language, developed by Microsoft, and version 3.5 of the Microsoft .NET Framework) with the ultimate specification of becoming being the primary interface platform for the realignment system. The interface has been implemented to the point of providing a simulated environment for the haptic device. It was found that the configuration of the RV-M1 provides a tight area of high dexterity as a haptic device, and as such, similar kinematic configurations are poor candidates for practical implementation. The implication of which, is that a new manipulator should be designed which grants a larger volume of high dexterity space.

Thesis (M.S.)--Auburn University, 2008. / Abstract. Vita. Includes bibliographical references (p. 70).

Ο σκοπός της διπλωματικής εργασίας ήταν η ανάπτυξη ενός μοντέλου Επαυξημένης Πραγματικότητας για την υποβοήθηση του χειρουργού-χειριστή ρομποτικού μηχανήματος. Το μοντέλο αυτό παρουσιάστηκε για την εύρεση του οπτικού πεδίου του ειδικού ενδοσκοπίου. Για τον σκοπό αυτό κατασκευάστηκαν δύο προγράμματα τα οποία μπορούν να χειριστούν ιατρικά δεδομένα και να προσφέρουν εικόνες από το εσωτερικό του μοντέλου του ασθενούς. Συγκεκριμένα, έγινε μελέτη των βασικών μεθόδων εφαρμογής Επαυξημένης Πραγματικότητας στην χειρουργική, όπως η εγγραφή του ασθενούς, η κατάτμηση των ιατρικών δεδομένων, η τρισδιάστατη ανακατασκευή τους και η ανίχνευση των ενδοσκοπικών εργαλείων και της κάμερας. Παρουσιάστηκε το πλήρες θεωρητικό μοντέλο εφαρμογής επαυξημένης πραγματικότητας και έγινε ανάλυση των επιμέρους διαδικασιών. Κατασκευάστηκαν με την βοήθεια της Matlab δύο προγράμματα με τα αντίστοιχα GUIs για τον προεγχειρητικό σχεδιασμό και την διεγχειρητική καθοδήγηση/επαύξηση αντίστοιχα. Τέλος έγινε δοκιμή των προγραμμάτων χρησιμοποιώντας 22 τομές μαγνητικής τομογραφίας (μορφής DICOM) εγκεφάλου με εμφανή καρκίνο στην αριστερή κοιλία. Επίσης καταγράφηκαν οι εικόνες και τα δεδομένα που παίρνουμε σε κάθε βήμα εφαρμογής των προγραμμάτων με στόχο την αξιολόγηση τους. Το μοντέλο αυτό κατασκευάστηκε με στόχο την εφαρμογή του σε επεμβάσεις μέσω του ρομποτικού μηχανήματος daVinci. Παρόλα αυτά η γενικότερη εφαρμογή της μεθοδολογίας που αναπτύσσεται μπορεί να βρει εφαρμογές και σε άλλες ενδοσκοπικές επεμβάσεις. / The purpose of this thesis was to develop a model of Augmented Reality to assist the surgeon-operator of a robotic machine. The model has been presented for finding the field of special endoscope. For this purpose we built two programs that can manipulate medical data and provide images of the interior of the patient’s model. Specifically, a study was done in the basic methods of Augmented Reality application in Surgery such as, the registration of the patient, the segmentation of medical data, their 3D reconstruction and the detection of endoscopic instruments and the camera. Has been presented the complete theoretical model for applying augmented reality and an analysis of individual procedures was done. Moreover we constructed with the help of Matlab two programs with their GUIs, for preoperative planning and intraoperative guidance/augmentation, respectively. Finally the programs were tested, using 22 MRI slices (format DICOM) with visible brain cancer in the left ventricle. Also were recorded images and data that we get at each step of programs implementation in order to evaluate them. This model was constructed to implement the operations through the daVinci robotic machine. Nevertheless, the general application of the methodology developed in this study may find applications also in other endoscopic procedures

Σκοπός της παρούσας διπλωματικής εργασίας είναι η ανάπτυξη και ο έλεγχος ρομποτικού συστήματος για την επεξεργασία βωξίτη. Αφορά συγκεκριμένα το μηχανολογικό σχεδιασμό του συστήματος, την επιλογή εξοπλισμού, τον ηλεκτρολογικό σχεδιασμό, τον έλεγχο με χρήση προγραμματιζόμενου λογικού ελεγκτή (PLC) και τέλος την υλοποίηση εφαρμογής scada / The purpose of this thesis is the development and control of robotic system for processing bauxite. Concern in particular the mechanical system design, the choice of equipment, the electrical design, the control using programmable logic controller (PLC) and finally the creation of scada application.

Resumo: A utilização do alumínio em escala industrial vem crescendo cada vez mais nos últimos anos e um dos principais motivos é o espaço que esse material vem conquistando em substituição ao aço. O interesse da indústria em ter um maior controle das distorções provocadas pela soldagem de alumínio se deve ao fato das tolerâncias dimensionais e geométricas estarem cada vez mais precisas nas especificações de projeto, motivando a engenharia de fabricação a desenvolver processos estáveis e que garantam a rotina de produção. O elevado coeficiente de expansão linear desse metal, se comparado ao aço, é uma das suas principais características que geram dificuldades nesse cenário. Visando isso, o presente trabalho tem como principal objetivo analisar o quanto situações rotineiras utilizadas na soldagem robotizada de alumínio podem influenciar nas distorções angulares deste material, sem comprometer a integridade mecânica da junta soldada e com confiabilidade estatística. Utilizando a liga AA 5052 H34, e a soldagem robotizada no processo GMAW pulsado, foram aplicados três tipos de tecimento ao longo do comprimento da solda, em juntas de topo montadas sem chanfro e com chanfro de 60º, dispostas tanto transversais quanto longitudinais ao sentido de conformação da chapa. A medição das distorções foi realizada por um braço tridimensional, antes e após a soldagem, em três regiões distintas nos corpos de prova. O perfil do cordão de solda foi o fator determinante para as diferentes formas das distorções encontradas, assim como revelado pelas análises macrográficas. A junta com chanfro de 60º apresentou maiores amplitudes das distorções que a junta sem chanfro. O tecimento não foi uma variável de influência estatística significativa nessas amplitudes. Ensaios de tração foram realizados para se avaliar o comportamento mecânico das juntas soldadas / Abstract: The use of aluminum by industry has been recently growing more and more each time and one of the main reasons is the position that this material is gaining as a replacement for steel. The industry's interest in having a greater control of the distorcions caused by aluminum welding is due to the geometric and dimensional tolerances been more and more precise in the project specifications, motivating the manufacturing engineering to develop stable processes and to ensure routine production. The high coefficient of linear expansion of this metal, which is one of its main features, can create a difficult scenario when compared to the steel. Aiming at it, the main goal of this present work is to analyze how much routine situations used int the robotic aluminum welding can influence in the angular distortions of this material, without affecting the mechanical integrity of the welded joint and with statistical reliability. Using the alloy AA 5052 H34, and the robotic welding in pulsed GMAW process, it was applied three types of weaving throughout the length of the weld, in butt joints assembled without groove and with 60º single-V-groove, arranged transversely as well as longitudinally to the rolling direction of the plate. The measurement of the distortions was made by a three-dimensional equipment, before and after the welding, in three distinct regions in the speciments. The profile of the weld bead was the main factor for the different types of distortions found, as revealed by macrographical analysis. The 60º single-V-groove had higher amplitudes of distortions as the joint without groove. The weaving wasn't a variable of statistically significant influence in this amplitudes. Tensile tests were made to evaluate the mechanical properties of the welded joints / Orientador: Yukio Kobayashi / Coorientador: Gilberto de Magalhães Bento Gonçalves / Banca: Américo Scotti / Banca: Luiz Eduardo de Angelo Sanchez / Mestre

La théorie du contrôle a d'abord été conçue pour des contrôleurs analogiques. Il était alors pertinent de synthétiser un contrôleur dans une représentation continue. De nos jours, les systèmes numériques ont majoritairement remplacés les systèmes analogiques pour différentes raisons (coût, résistance au bruit, intégration,...). Les signaux sont alors maintenus constants entre le cadencement périodique fixé par une horloge numérique.La commande événementielle vise à améliorer l’échantillonnage périodique en proposant une méthode dans laquelle les mises à jour sont déclenchées par une fonction événement.Dans cette thèse, de nouvelles méthodes de synthèse de contrôleurs evenementiels sont présentés et testés sur des systèmes temps-réels.La contribution la plus originale étant l'utilisation d'une commande événementielle appliquée à un problème d'anticollision entre une quadrirotor et un environnement supposé connu. / Control system theory has first been built for analog controllers. In this context, it was relevant to use a continuous framework to design a control feedback function. Nowadays, digital technologies are supplanting analog solutions due to several advantages (cost, noise, resistance, integration...). Signal is held constant between periodic triggers given by a digital clock.Event-based (or event-triggered) control aims to improve the periodic sampling scheme by proposing a method in which updates are triggered by an event function.In this thesis new event-based designs are detailled and tested on real-time systems.The most original contribution of this thesis is the use of an event-triggered design for a problem of collision avoidance between a quadrotor and a known environment.

[pt] O método da dinâmica inversa para o controle de manipuladores robóticos é apresentado. A ideia básica deste método é cancelar as não linearidades e acoplamentos, que caracterizam o comportamento dinâmico de manipuladores, através de um modelo dinâmico do mesmo (controlador primário). Com isto, o sistema resultante é linear e desaclopado, podendo ser controlado por técnicas de controle linear (controlador secundário). O método é inicialmente desenvolvido considerando o caso ideal do controlador primário (onde o modelo dinâmico é perfeito) e um PD para o controlador secundário. As implicações de imperfeições no cancelamento das não linearidades e aclopamentos do sistema pelo controlador primário são mostradas. As duas formulações existentes para o controlador primário – computed – torque e feedforward – são descritas. É sugerida uma formulação híbrida para contornar os problemas de implementação das duas formulações. Um enfoque maior é dado às versões simplificadas da formulação computed – torque. Simulações são feitas para melhor esclarecer esta questão. Em substituição ao PD, é descrito o projeto de um compensador linear robusto a partir do método das fatorações por matrizes própias e estáveis. O projeto é apresentado com análises mais detalhadas de algumas questões e com correções nos erros encontrados, em relação ao projeto existente na literatura. Análises comparativas com o PD são feitas e é explicada a influencia de frequências de amostragem no desempenho e ganhos do controlador PD. / [en] The inverse dynamics control of robot manipulators is presented. The main idea of this control method is to cancel the nonlinearities and coupling effects, that describe the dynamic behavior of manipulators, using a dynamic model of the system (primary controller). Since the resulting system is linear and uncoupled, it can be controlled by linear control techniques (secondary controller). The method is initially derived considering the ideal case of the primary controller (where the dynamic model is perfect) and a PD for the secondary controller. The implications of inexact cancelling of the system nonlinearities and coupling effects by the primary controller are shown. The two existing primary controller formulations – computed-torque and feedforward – are described. A hybrid formulations is suggested to overcome the implementation problems of the two formulations. Special attention is given to the simplified computed-torque schemes, which are subject of controversy in the literature. Simulations are performed to better illustrate this matter. A robust linear compensator design, based on the stable factorization approach, is described analyses of some questions and with corrections of the detected mistakes, regarding the existing design. Comparative analyses with the PD are done. The effects of sampling rates on the tracking performances and PD gains are explained.

Depuis maintenant près d’un siècle, des robots sous-marins ont été développés afin de réaliser des tâches spécifiques aux besoins des grands acteurs historiques du domaine (militaires, pétroliers, câbliers ou explorateurs benthiques) sans vraiment se soucier de la polyvalence et de la modularité de la plateforme. L’objectif de cette thèse est d’avoir une réflexion sur une solution technologique et scientifique avec comme domaine applicatif l’environnement faible fond ou confiné. Il s’agit en d’autres termes de concevoir un robot sous-marin que l’on peut faire évoluer aisément tant sur le plan mécanique, électronique qu’informatique. Cet objectif impose de proprement conceptualiser cette « polyvalence » en s’attachant à apporter de l’abstraction dans l’architecture de contrôle que se soit au niveau de l’automatique avec l’expression de la polyvalence liée à l’étage d’actionnement ou de l’informatique avec une architecture basée services, pouvant être composés, afin de répondre à la diversité des besoins applicatifs. L’ensemble de ces travaux a pour point de départ le robot Jack mis au point par l’entreprise Ciscrea, partenaire industriel, qui apporte un aspect économique à la nécessité de développer une solution polyvalente pouvant être décliné en une gamme de produits. Ce manuscrit traite de la conception, la réalisation et l’expérimentation de ce vecteur que ce soit en piscine où en environnement réel. / Since almost a century, underwater robots have been developed in order to respond to the specific needs of historical actors of the domain (military, hydrocarbons exploitation, underwater cabling or benthic exploration), without addressing specifically the question of versatility or modularity of the underwater platform. This thesis aims to address these questions on a technological solution dedicated to shallow water or confined environment. In other words, the objective is to realise an underwater system, able to evolve on the mechanical, electronical or software aspects. This requires to properly conceptualise this « versatility » with an abstraction of the control architecture, on the actuation aspect, with the expression of the versatility linked to the actuation systems, or on the software architecture level, with a Service-Oriented-Architecture (SOA) approach, in order to tackle the diversity of the application requirements. This study is based on the Jack system, developed by the Ciscrea Company, which is the industrial partner of this project, and brings the economical aspect as a central requirement. This underlines another view of the versatility question, the development of a range of product for the Ciscrea Company.This thesis proposes the conception, realisation and experimentation of such a versatile underwater system, with test-tank and field validation.

BACKGROUND: Hysterectomies are one of the most frequently performed surgical procedures in the United States. There are a wide variety of diagnoses that require a patient to obtain this procedure, but the majority of hysterectomies are performed for benign indications. Currently, gynecologists do not follow a standardized protocol surrounding postoperative laboratory ordering, and healthcare professionals can order a wide range of tests as often as they choose. Extraneous laboratory orders are disruptive to the patients’ well-being and risk their health following surgery. These orders are costly for hospital systems, take up precious time of hospital employees, and influence the course of patient treatment only in extremely rare circumstances. There are few studies that develop exclusion criteria for patients who may not require a laboratory test following surgery. Though systems to predict postoperative hematocrit have been created, they are complicated and difficult to use. The few studies that were performed are yet to be accepted by the medical community, in part because of their limited scope. This study will be the first to incorporate the results of robotic surgery in the analysis. OBJECTIVE: The purpose of this study is to determine concrete parameters to indicate that a patient is in need of postoperative laboratory work and at risk for anemia or transfusion. We aim to develop two comprehensive models that guide surgical practitioners to identify the cases which do not require laboratory data. METHODS: A total of 1027 gynecologic surgeries were performed at Saint Francis Hospital and Medical Center between April 1, 2014 and May 31, 2016. This retrospective study extracted data from EPIC EMR according to 42 variables preconceived to be the leading indicators of postoperative hematocrit and overall healing. Five healthcare professionals were surveyed to identify the variables that influence their postsurgical patient assessments and their decisions to order blood testing. This information was developed into score sheets with differing levels of stringency. Correlation highlighted 14 of the initial 42 variables as contributors to postoperative hematocrit and an equation model was built. Stepwise linear regression was used for univariate and multivariate analyses, from which we created our equation to predict all patients’ postoperative hematocrit. RESULTS: Out of the 1027 initial cases, a total of 602 cases were identified as hysterectomies for benign indications. Survey data gave the highest value to urine output and heart rate as key indicators of postoperative anemia. From the survey data, two clinical scoring sheets with differing stringency were created to guide practitioner laboratory ordering. These sheets gave parameters of heart rate and urine output the largest correlative weight in determining postoperative hematocrit. However, based on regression analysis, parameters of age (AGE), body mass index (BMI), preoperative platelet count (PPC), estimated blood loss during surgery (IO EBL), preoperative hematocrit (PHCT) and postoperative fluid bolus orders (POSTOP FB) proved to be the key variables impacting postoperative hematocrit (POSTOP HCT). These items were translated into the equation: POSTOP HCT = 22.51 – 0.40*POSTOP FB – 0.01*IO EBL + 0.25 PHCT + 0.09*BMI + 0.06*AGE – 0.01*PPC (R-squared = 0.310). CONCLUSIONS: This study aims to decrease superfluous laboratory testing, as well as to contribute to a larger conversation considering the potential merits of clinical judgement in a data-driven healthcare system. We have created a number of comparable strategies in order to reduce the number of unnecessary blood draws: two clinical scoring sheets and an equation. The score sheets indicate when to order additional testing. These sheets are representative of a range of surgical practitioners’ conventional clinical judgement. The equation serves as an evidence-based guide for determining postoperative hematocrit following benign gynecologic surgery. These predictive mechanisms will be validated and a superior method determined as our research continues with prospective application. We eventually expect to use the most accurate mechanism to reduce postoperative blood testing following all surgeries.