"The use of cameras and computer vision algorithms to provide state estimation for robotic systems has become increasingly popular, particularly for small mobile robots and unmanned aerial vehicles (UAVs). These algorithms extract information from the camera images and perform simultaneous localization and mapping (SLAM) to provide state estimation for path planning, obstacle avoidance, or 3D reconstruction of the environment. High resolution cameras have become inexpensive and are a lightweight and smaller alternative to laser scanners. UAVs often have monocular camera or stereo camera setups since payload and size impose the greatest restrictions on their flight time and maneuverability. This thesis explores ORB-SLAM, a popular Visual SLAM method that is appropriate for UAVs. Visual SLAM is computationally expensive and normally offloaded to computers in research environments. However, large UAVs with greater payload capacity may carry the necessary hardware for performing the algorithms. The inclusion of general-purpose GPUs on many of the newer single board computers allows for the potential of GPU-accelerated computation within a small board profile. For this reason, an NVidia Jetson board containing an NVidia Pascal GPU was used. CUDA, NVidia’s parallel computing platform, was used to accelerate monocular ORB-SLAM, achieving onboard Visual SLAM on a small UAV. Committee members:"

In the information age, intelligent indoor positioning and navigation services are required in many application scenarios. However, most current visual positioning systems cannot function alone and have to rely on additional information from other modules. Nowadays, public places are usually equipped with monitoring cameras, which can be exploited as anchors for positioning, thus enabling the vision module to work independently. In this thesis, a high-precision indoor positioning and navigation system is proposed, which integrates monitoring cameras and smartphone cameras. Firstly, based on feature matching and geometric relationships, the system obtains the transformation scale from relative lengths in the cameras’ perspective to actual distances in the floor plan. Secondly, by scale transformation, projection, rotation and translation, the user's initial position in the real environment can be determined. Then, as the user moves forward, the system continues to track and provide correct navigation prompts. The designed system is implemented and tested in different application scenarios. It is proved that our system achieves a positioning accuracy of 0.46m and a successful navigation rate of 90.6%, which outperforms the state-of-the-art schemes by 13% and 3% respectively. Moreover, the system latency is only 0.2s, which meets the real-time demands. In summary, assisted by widely deployed monitoring cameras, our system can provide users with accurate and reliable indoor positioning and navigation services. / Thesis / Master of Applied Science (MASc)

La tâche consistant à tirer de l'information des images d'une caméra au cours du temps pour cartographier l'environnement et se localiser à l'intérieur de celui-ci, est appelée Localisation et Cartographie Simultanée ou SLAM.Développée à la fois par les communautés scientifiques de robotique et vision par ordinateur les applications sont multiples. Des robots bénéficient de cette capacité en gagnant en autonomie. Ces dernières années, des résultats impressionnants ont été obtenus pour des applications à des moyens de transport autonomes.Une autre champ d'application est la réalité augmenté. La localisation donnée par le SLAM offre la possibilité d'obtenir un rendu des éléments virtuels en cohérence avec les mouvements de l'utilisateur. Ainsi le cinéma, les jeux vidéos, le tourisme peuvent bénéficier de techniques SLAM. L'assistance aux travailleurs effectuant des tâches de précision ou répétitives compte également parmi les champs d'application du SLAM. Dans le cadre de cette thèse nous nous sommes intéressés au SLAM dans une optique d'applications réalistes de réalité augmentée. Bien que le sujet ait été beaucoup exploré et que d'intéressants résultats aient été obtenus, la tâche n'est toujours pas parfaitement résolue. Le problème du SLAM est un sujet de recherche ouvert, aussi bien sur des aspects spatiaux (dérive, fermeture de boucle) que temporels (temps de traitement). Dans le cadre du SLAM monoculaire nous avons surtout adressé le problème de la dérive. Puis nous nous sommes intéressés au SLAM multi-capteurs, afin d'adresser le problème des mouvements de rotation problématiques dans le cas monoculaire, et celui de la complexité calculatoire. / Gathering informations from the images of a camera, over time, in order to map the environment and localize the camera in it, is a task refered to as Simultaneous Localization and Mapping, or SLAM. Developped both by the robotics and computer vision scientific communities, its applications are many. Robots gain autonomy from this ability. Quite recently, impressive results have been obtained in applications to autonomous transportation vehicles. Another field of application is augmented reality. The localization offered by SLAM enables us to display virtual objects in a consistent way a user movements. Thus, cinema, video games, tourisme applications can benefit from SLAM methods. Visual aids to workers performing complex or repetetive tasks is also an interesting application of SLAM methods. During this PhD thesis, we took interest in SLAM with the idea of realistic augmented reality applications in mind. Though the topic has been extensively explored and many impressive results obtained, the task isn't completely solved. The problem is still an open one, regarding spatial facets (drift, loop closure) as well as temporal (processing time). As part of our monocular SLAM explorations, we mainly studied the drift issue. We then explored multisensor SLAM, both as a mean to handle problematical rotational movements for the monocular setup and as mean to reduce the substantial processing times needed to solve the problem.

Cette thèse traite du problème de localisation et contrôle de robots humanoïdes par rapport à leur environnement, tel qu'observé par ses capteurs embarqués. Le SLAM visuel dense, consistant en l'estimation simultanée d'une carte 3D de l'environnement et de la position du robot dans cette carte est exploité pour étendre et robustifier les méthodes de planification contrôle multi-contact. Celles-ci consistent à établir et exploiter des contacts robot-environnement pour accomplir des tâches de locomotion et manipulation. Des incertitudes sur la posture initiale du robot, ainsi que des perturbations causées par une modélisation inadéquate des contacts, ainsi que des perturbations externes oblige à la prise en compte de l'état du robot et son environnement. Une méthode de calibration corps-complet est également proposée, afin d'obtenir une connaissance fiable de la chaîne cinématique du robot, nécessaire pour réaliser de telles tâches. Finalement, une méthode de marche basée sur de la commande prédictive de modèles est robustifiée par la prise en compte de large perturbations, permettant d'ajuster les trajectoires de pied et du centre de masse afin de garantir sa stabilité, tout en accomplissant les objectifs désirés. Les méthodes proposées sont illustrées et validées par de multiples expérimentations sur les robots humanoïdes HRP-2Kai et HRP-4. / This thesis deals with the problem of localizing and controlling humanoid robots with respect to its environment, as observed by its on-board sensors. Dense visual SLAM, consisting in the simultaneous estimation of a 3D map of the environment and of the robot localization within that maps is exploited to extend and robustify multi-contact planning and control. Establishing and exploiting robot-environment contacts allows the accomplishment of both locomotion and manipulation tasks. Uncertainties in the initial robot posture, and perturbations arising from improper contact-modelling and external causes are accounted for by observing the state of the robot and its environment. A whole-body calibration method is also proposed, so that robust knowledge of the robot's kinematic structure is known, a prerequisite to all robot-environment interaction tasks. Finally, a walking method based on model predictive control is robustified by taking into account large perturbations, and adjusting the footstep and center-of-mass trajectories accordingly to guarantee stability while accomplishing desired objectives.Several experiments on an HRP-2Kai and an HRP-4 humanoid robots are presented and discussed to illustrate and validate each of the proposed methods.

An on-line 3D visual object tracking framework for monocular cameras by incorporating spatial knowledge and uncertainty from semantic mapping along with high frequency measurements from visual odometry is presented. Using a combination of vision and odometry that are tightly integrated we can increase the overall performance of object based tracking for semantic mapping. We present a framework for integration of the two data-sources into a coherent framework through uncertainty based fusion/arbitration.

Sammanfattning Totalt 80 % av världens energianvändning kommer från förbränning av fossila bränslen. Användningen tros också öka med tiden vilket skulle påverka den förstärkta växthuseffekten ytterligare. De ökade utsläppen av växthusgaser påverkar klimatförändringarna och är någonting som alla borde bidra till att minska. I Sverige är det mindre än 30 % av den totala energianvändningen som kommer från fossila bränslen. Det är dock inte tillräckligt för en hållbar utveckling.   Tillverkningen av papper och massa förbrukar stora mängder vatten som senare måste renas för att kunna släppas ut. Vid reningen bildas slam som ofta förbränns. Istället för att förbränna slammet går det att ta till vara på energin genom att röta slammet och få ut biogas. Dock är det svårt att få detta ekonomiskt hållbart. Studier visar att med hjälp av förbehandling av bioslammet kan större mängder biogas produceras och därmed göra processen mer lönsam. Biogasen som bildas genom rötning kan sedan ersätta en viss del av de fossila bränslen som används i olika processteg i bruken och därmed föra pappers- och massaindustrin ett steg närmare ett hållbart koncept.   Förbehandlingar som visats gynna biogasproduktionen är kombination av kemisk och termisk behandling som sönderdelar slammet innan rötningen. Då kan biogasproduktionen öka och mer fossila bränslen kan fasas ut.   I studien behandlas bioslam kemiskt och termiskt innan rötning. De kemikalier som undersöktes var natriumhydroxid (NaOH), kaliumhydroxid (KOH) och kalciumhydroxid Ca(OH)2 som användes för att ge slammet pH 10 eller 12. De termiska förbehandlingarna utfördes i 100 °C respektive 140 °C. Som referenspunkter användes obehandlat slam och endast värmebehandlat slam. Bioslammet som användes hämtades på Stora Ensos bruk i Skoghall och det kommunala rötslammet från Fiskaretorpets reningsverk i Kristinehamn då de använder mesofil rötning.   Rötningen har skett satsvis med hjälp av en AMPTS2 (ett analytiskt instrument för rötningsprocesser) i tre rötningsomgångar. Den förbehandling som gav allra mest metangas var NaOH pH 12 i 100 °C. Detta skedde i rötningsomgång två. I de andra rötningsomgångarna var det förbehandlingen med NaOH pH 12 i 140 °C respektive KOH pH 12 i 140 °C som genererade mest metangas. Vid jämförelse av de olika rötningsomgångarna användes rötning av cellulosa för respektive omgång som referens. Skillnaderna mellan de tre som producerade mest metangas var små. Förbehandling med pH 12 gav mer metangas än med pH 10. Förbehandlingar med NaOH producerade mer metangas vid 100°C än vid 140 °C, medan det motsatta gällde för förbehandlingarna med Ca(OH)2. / Abstract A total of 80 % of the world’s energy use comes from fossil fuels, which causes emissions of greenhouse gases leading to climate change. In Sweden less than 30 % of the total energy use comes from fossil fuels. It must however, be decreased further to achieve a sustainable development.   A large amount of water is used during the manufacturing of pulp and paper. The water has to be treated before reaching a recipient and during this purification process, sludge is left as a by-product. The sludge is usually incinerated, although it could also be used in an anaerobic digestion process to produce biogas. The method is difficult to make economically viable, but studies show that by pre-treatment of the sludge, larger amounts of biogas can be produced, which would make the process more profitable. Biogas produced by anaerobic digestion can replace fossil fuels in some processes at pulp and paper mills and therefore contribute to a smaller environmental footprint for this industry.   Pre-treatments shown to promote the production of biogas are a combination of chemical and thermal treatment, which decomposes the sludge before digestion. The biogas production will then increase and more fossil fuels can be phased out.   In this report, biosludge is treated chemically and thermally before digestion. The chemicals tested were sodium hydroxide, potassium hydroxide and calcium hydroxide, added until the sludge had a pH of 10 or 12. The thermal treatments tested were 100 °C and 140 °C. As references, untreated sludge and only heat treated sludge were used. The biosludge that was used was collected at Stora Enso CTMP-mill in Skoghall and the inoculum were collected from a municipal treatment plant.    Digestion was done in batch with the help of an AMPTS2 in three digestion rounds. The pre-treatment that produced the most methane was 100 °C with NaOH at pH 12. This was in anaerobic digestion round two. In the other digestion rounds the best pre-treatments were NaOH pH 12 at 140 °C and KOH pH 12 at 140 °C. When comparing the different rounds, a cellulose sample was used for each round as a reference. All chemical pre-treatments generated more methane at pH 12 than at pH 10. Pre-treatments with NaOH produced more methane gas at 100°C than at 140 °C, whereas the opposite was true for Ca(OH)2.

Magíster en Ciencias de la Ingeniería, Mención Ingeniería Eléctrica / Ingeniero Civil Eléctrico / El uso de Conjuntos Aleatorios Finitos (RFS por su sigla en inglés) tiene varias ventajas respecto de los métodos tradicionales basados en vectores. Entre ellas están el incluir las estadísticas de detección del sensor y la eliminación de las heurísticas tanto para la asociación de datos como para la inicialización y eliminación de objetos en mapa. Para obtener los beneficios de los estimadores basados en RFS en el problema de Construcción de Mapas y Localización Simultanea (SLAM por su acrónimo en inglés), las estadísticas de detección y falsa alarma del extractor de características deben ser modeladas y utilizadas en cada actualización del mapa. Esta Tesis presenta técnicas para obtener estas estadísticas en el caso de características semánticas extraídas de mediciones láser. Además se concentra en la extracción de objetos cilíndricos, como pilares, árboles y postes de luz, en ambientes exteriores. Las estadísticas de detección obtenidas son utilizadas dentro de una solución a SLAM basada en RFS, conocida como Rao-Blackwellized (RB)-probability hypothesis density (PHD)-SLAM, y el algoritmo multiple hypothesis (MH)-factored solution to SLAM (FastSLAM), solución a SLAM basada en vectores. El desempeño de cada algoritmo al usar estas estadísticas es comparado con el de utilizar estadísticas constantes. Los resultados muestran las ventajas de modelar las estadísticas de detección, particularmente en el caso del paradigma RFS. En particular, el error en las estimaciones del mapa, medido utilizando la distancia optimal sub- pattern assignment (OSPA) a un mapa ground truth generado de forma independiente, disminuye en un 13% en el caso de MH-FastSLAM y en un 13% para RB-PHD-SLAM al modelar las estadísticas de detección. A pesar de que no se tiene un ground truth para la trayectoria del robot, se evalúan las trayectorias visualmente, encontradose estimaciones superiores para el método propuesto. Por lo tanto, se concluye que el modelamiento de las estadísticas de detección es de gran importancia al implementar una aplicación de SLAM.

Syftet för studien är att ta fram möjliga avsättningsalternativ för slammet från Duvbackens avloppsreningsverk i Gävle och med stöd av en multikriterieanalys bedöma vilket alternativ som anses vara lämpligast med avseende på teknisk tillgänglighet, miljömässig hållbarhet, ekonomiskt perspektiv samt hälsa och hygien. Tre metoder har använts för studien; litteraturstudie, personlig kontakt samt multikriterieanalys. Duvbackens slam uppfyller inte dagens krav för att spridas på jordbruksmark och inte heller kraven för mindre känslig markanvändning uppnås, därmed måste andra alternativ utvärderas. De tre avsättningsalternativ som kan föreslås för slammet från Duvbacken, förutom nollalternativet är; täckningsmaterial till deponi, förbränning i kraftvärmeverk, gödsel till skogsmark och möjligen gödsel till salixodling (energiskog) det sistnämnda är möjligt endast om dispens från Länsstyrelsen kan erhållas. Dessa fyra alternativ utvärderas i en multikriterieanalys, utifrån viktningsgruppens prioriteringar. Medelvärdet för gruppens prioriteringar visar till exempel att de kriterier som anses viktiga är bland annat flexibilitet, resursutnyttjande, efterfrågan samt arbetsmiljö. De slutliga betygen blev relativt jämt fördelade för de olika alternativen, men salix fick högst slutbetyg, därefter skogsbruk, täckningsmaterial, och slutligen förbränning som fick lägst slutbetyg. Alla avsättningsalternativ har för- och nackdelar och spridning av de oönskade ämnena går inte att undkomma så länge de uppkommer från första början, därför är den bästa satsningen uppströmsarbete, e.g. att förebygga att dessa ämnen hamnar i avloppet vid källan. Slutsatser som kan dras av denna studie är att salixodling anses som det lämpligaste avsättningsalternativet för Duvbackens slam enligt utförd multikriterieanalys. I ett långsiktigt helhetsperspektiv är dock uppströmsarbete troligen den bästa investeringen. / The aim of the study is to provide and evaluate possible disposal alternatives for sludge from the wastewater treatment plant Duvbacken in Gävle by using a multicriteria analysis with respect to technical availability, environmental sustainability, economic perspective and health & hygiene. Three methods have been used in the study; literature study, personal contact and a multicriteria analysis. The sewage sludge quality today at Duvbacken does not meet national requirements to be used in agricultural purposes nor within less sensitive land use. Thus, other options should be investigated. Three deposal options have been suggested in this study by considering the legal basis and sludge quality, these are; combustion in combined heat and power plant, fertilizer to forest and fertilizer to willow (salix) cultivation, the latter is only possible if an exemption from the provincial government can be obtained. Those three alternatives, together with current landfill deposition, were evaluated in a multicriteria analysis, according to the weighting group`s priorities. The average value of the group´s priorities shows for example that important criteria seem to be flexibility, resource utilization, demand and work environment. The final results was relatively even among the options, willow got the highest rating, then forestry, then landfill cover, combustion received the lowest score. All options have pros and cons and the spreading of unwanted substances cannot be completely avoided as long as they exist in the incoming flows into the wastewater treatment plant, therefore the best investment might be upstream investigation, e.g. preventing these substances at their sources. Conclusion from this study is that willow cultivation seems to be the best disposal option for sludge from Duvbacken according to the multicriteria analysis. However, in a sustainable holistic perspective, upstream investigation would probably be the best investment.

No description available.

The problem of robotic mapping, also known as simultaneous localization and mapping (SLAM), by a mobile agent for large distributed environments is addressed in this dissertation. This has sometimes been referred to as the holy grail in the robotics community, and is the stepping stone towards making a robot completely autonomous. A hybrid solution to the SLAM problem is proposed based on "first localize then map" principle. It is provably consistent and has great potential for real time application. It provides significant improvements over state-of-the-art Bayesian approaches by reducing the computational complexity of the SLAM problem without sacrificing consistency. The localization is achieved using a feature based extended Kalman filter (EKF) which utilizes a sparse set of reliable features. The common issues of data association, loop closure and computational cost of EKF based methods are kept tractable owing to the sparsity of the feature set. A novel frequentist mapping technique is proposed for estimating the dense part of the environment using the sensor observations. Given the pose estimate of the robot, this technique can consistently map the surrounding environment. The technique has linear time complexity in map components and for the case of bounded sensor noise, it is shown that the frequentist mapping technique has constant time complexity which makes it capable of estimating large distributed environments in real time. The frequentist mapping technique is a stochastic approximation algorithm and is shown to converge to the true map probabilities almost surely. The Hybrid SLAM software is developed in the C-language and is capable of handling real experimental data as well as simulations. The Hybrid SLAM technique is shown to perform well in simulations, experiments with an iRobot Create, and on standard datasets from the Robotics Data Set Repository, known as Radish. It is demonstrated that the Hybrid SLAM technique can successfully map large complex data sets in an order of magnitude less time than the time taken by the robot to acquire the data. It has low system requirements and has the potential to run on-board a robot to estimate large distributed environments in real time.