Search

State estimation is a critical component of an autonomous vehicle. Safe, accurate and reliable localization can be a challenging task, especially in worst-case scenarios where sensors may suffer from poor performance or failure (for instance, a GPS system inside a building). In these cases, the sensors provide corrupt measurements. Previous works have studied the fusion of valid information from different sensors, however the sensor rejection problem has barely been explored in a general, non algorithm-specific formulation. In this thesis, the online cross-validation and rejection of pose estimates coming from sensors that may fail is studied. This is the applied to the Autonomous Plus2, an autonomous vehicle prototype for last-mile delivery developed by Kyburz Switzerland AG. First, a general overview of autonomous vehicle sensors and their typical failure modes is presented. The specific localization algorithms present on the Autonomous Plus2 are also described. Then, existing sensor rejection approaches are explored, based on parity-based Fault Detection and Isolation (FDI) techniques. New approaches are studied, based on Machine Learning (ML) techniques (decision trees, k-NN, feedforward neural networks and RNNs). A solution combining the aforementioned techniques is proposed, to improve reliability. Finally, an experimental setup is presented and implemented to test and tune each algorithm on the Autonomous Plus2. Various challenging scenarios are considered, to assess performance, trade-offs and weaknesses. The final proposal is a combined decision algorithm, capable of a 70% rejection rate of failed measurements while only losing 5% of the valid measurements., Objectius de Desenvolupament Sostenible::9 - Indústria, Innovació i Infraestructura

Walking has multiple health benefits. One way to increase students’ walking activity is by encouraging active morning commutes. However, students may compensate for active commutes by walking less throughout the day, rendering such initiatives ineffective in increasing overall walking activity. This study aims to assess how morning commuting modes affect students’ walking levels, hypothesizing that gains in walking from active morning commutes may not be sustained throughout the day due to compensatory behavior. Our study analyzed objectively measured, sensor-collected data of 5600 children (ages 7 to 18) in Singapore for up to four consecutive weekdays between September and November 2015. Potential confounders of age, socioeconomic status, and built environment characteristics, as well as home-school distance as an effect modifier, were examined. We used linear mixed effects models to analyze differences in step count between students with different morning modes, as well as to analyze ‘within students’ variations when students switched between different modes over different days. Students who walked or took public transport walked more than their driven peers during morning commuting hours, by 96.1 steps per hour (95% CI = 71.5, 120.8) and 54.1 steps per hour (95% CI = 32.2, 75.9) respectively. Students who switched morning commute modes from car to public transport took 47.6 more morning steps per hour (95% CI = 10.3, 84.9) when using public transport, compared to when driven. However, the relationship between morning travel modes and step count per hour across the full day was less clear-cut. Both our ‘between students’ and ‘within students’ analyses suggest that taking more active morning modes, after controlling for all possible confounders and modifers, was not associated with higher step counts over the entire day. Encouraging students to walk more through more active morning commutes alone may have limited effectiveness in increasing overall daily walking a

In recent years vehicle powertrains show a growing trend towards diversification. This helps to mitigate inevitable conflicts between objectives, but vastly expands and complicates the concept solution space. The fair optimization and assessment across a number of distinct powertrain structures poses a considerable challenge for vehicle concept teams in the early development phase. Previous work fails to derive a realistic and holistic approach to tackle this challenge, which is still viable under tight time constraints. The solution presented in this thesis is based on novel combination of state-of-the-art concepts. Flexible powertrain definition is achieved with graph theory and parametric component models. Important secondary mass decompounding effects are considered. A vectorized quasi-static backward simulation approach and derived meta-models guarantee high computational performance. Fair dynamic assessment is realized consistently with global optimal control. Component sizing uses closed-loop optimization. The implementation of the methodology is applied for five technical vehicle criteria and to 10 powertrain structures as an example. For each structure two criteria are simultaneously optimized and depicted as a Pareto front. Concurrent visualization of the respective components' characteristics provides transparent insight on the relation between objective and decision space. Families of Pareto fronts allow comparative graphical decision making. The methodology presented here contributes to advanced vehicle systems engineering and its application in a practical development context., In den letzten Jahren sind Fahrzeugantriebsstränge durch einen Trend zur Diversifizierung geprägt. Dies hilft Zielkonflikte zu entschärfen, führt jedoch zu einem wesentlich größeren und komplexeren Konzeptlösungsraum. In der frühen Phase stellt die faire Optimierung und Bewertung von verschiedenen Alternativen eine wesentliche Herausforderung für Fahrzeugkonzeptteams dar. Bisherige Arbeiten bieten keinen realistischen und ganzheitlichen Ansatz zur Lösung dieser Herausforderung, der auch im Rahmen von straffen Zeitvorgaben anwendbar ist. Die Methode in dieser Arbeit basiert auf der neuartigen Kombination von bestehenden Konzepten. Flexible Definition von Antriebstrangstrukturen wird erreicht durch Graphentheorie und parametrische Komponentenmodelle. Wichtige sekundäre Masseneffekte werden berücksichtigt. Vektorisierte quasi-statische Rückwärtssimulation und abgeleitete Meta-Modelle sorgen für eine hohe Recheneffizienz. Faire dynamische Bewertung wird konsistent durch global-optimale Antriebsstrangregelung sichergestellt. Die Komponentendimensionierung erfolgt mit einer Optimierungsschleife. Die Implementierung der Methode wird beispielhaft für fünf technische Bewertungskriterien und auf 10 Antriebsstrangstrukturen angewendet. Für jede Struktur werden zwei Kriterien gleichzeitig optimiert und als Pareto-Front dargestellt. Transparenz in der Beziehung zwischen Bewertungs- und Entscheidungsraum ermöglicht die simultane Visualisierung mit den entsprechenden Komponentencharakteristika. Die Kombination zu Pareto-Front-Scharen ermöglicht eine graphisch vergleichende Bewertung und Entscheidung zwischen optimierten Antriebsstrangstrukturen. Die in dieser Arbeit entwickelte Methode ist ein Beitrag zur verbesserten Systemauslegung von Fahrzeugen und ihrer Anwendung in einem praktischen Entwicklungsumfeld.

In recent years vehicle powertrains show a growing trend towards diversification. This helps to mitigate inevitable conflicts between objectives, but vastly expands and complicates the concept solution space. The fair optimization and assessment across a number of distinct powertrain structures poses a considerable challenge for vehicle concept teams in the early development phase. Previous work fails to derive a realistic and holistic approach to tackle this challenge, which is still viable under tight time constraints. The solution presented in this thesis is based on novel combination of state-of-the-art concepts. Flexible powertrain definition is achieved with graph theory and parametric component models. Important secondary mass decompounding effects are considered. A vectorized quasi-static backward simulation approach and derived meta-models guarantee high computational performance. Fair dynamic assessment is realized consistently with global optimal control. Component sizing uses closed-loop optimization. The implementation of the methodology is applied for five technical vehicle criteria and to 10 powertrain structures as an example. For each structure two criteria are simultaneously optimized and depicted as a Pareto front. Concurrent visualization of the respective components' characteristics provides transparent insight on the relation between objective and decision space. Families of Pareto fronts allow comparative graphical decision making. The methodology presented here contributes to advanced vehicle systems engineering and its application in a practical development context., In den letzten Jahren sind Fahrzeugantriebsstränge durch einen Trend zur Diversifizierung geprägt. Dies hilft Zielkonflikte zu entschärfen, führt jedoch zu einem wesentlich größeren und komplexeren Konzeptlösungsraum. In der frühen Phase stellt die faire Optimierung und Bewertung von verschiedenen Alternativen eine wesentliche Herausforderung für Fahrzeugkonzeptteams dar. Bisherige Arbeiten bieten keinen realistischen und ganzheitlichen Ansatz zur Lösung dieser Herausforderung, der auch im Rahmen von straffen Zeitvorgaben anwendbar ist. Die Methode in dieser Arbeit basiert auf der neuartigen Kombination von bestehenden Konzepten. Flexible Definition von Antriebstrangstrukturen wird erreicht durch Graphentheorie und parametrische Komponentenmodelle. Wichtige sekundäre Masseneffekte werden berücksichtigt. Vektorisierte quasi-statische Rückwärtssimulation und abgeleitete Meta-Modelle sorgen für eine hohe Recheneffizienz. Faire dynamische Bewertung wird konsistent durch global-optimale Antriebsstrangregelung sichergestellt. Die Komponentendimensionierung erfolgt mit einer Optimierungsschleife. Die Implementierung der Methode wird beispielhaft für fünf technische Bewertungskriterien und auf 10 Antriebsstrangstrukturen angewendet. Für jede Struktur werden zwei Kriterien gleichzeitig optimiert und als Pareto-Front dargestellt. Transparenz in der Beziehung zwischen Bewertungs- und Entscheidungsraum ermöglicht die simultane Visualisierung mit den entsprechenden Komponentencharakteristika. Die Kombination zu Pareto-Front-Scharen ermöglicht eine graphisch vergleichende Bewertung und Entscheidung zwischen optimierten Antriebsstrangstrukturen. Die in dieser Arbeit entwickelte Methode ist ein Beitrag zur verbesserten Systemauslegung von Fahrzeugen und ihrer Anwendung in einem praktischen Entwicklungsumfeld.

OBJECTIVE: In the human pathogen, Staphylococcus aureus, the agr quorum sensing system controls expression of a multitude of virulence factors and yet, agr negative cells frequently arise both in the laboratory and in some infections. The aim of this study was to examine the possible reasons behind this phenomenon.RESULTS: We examined viability of wild type and agr mutant cell cultures using a live-dead stain and observed that in stationary phase, 3% of the wild type population became non-viable whereas for agr mutant cells non-viable cells were barely detectable. The effect appears to be mediated by RNAIII, the effector molecule of agr, as ectopic overexpression of RNAIII resulted in 60% of the population becoming non-viable. This effect was not due to toxicity from delta toxin that is encoded by the hld gene located within RNAIII as hld overexpression did not cause cell death. Importantly, lysed S. aureus cells promoted bacterial growth. Our data suggest that RNAIII mediated cell death of agr positive but not agr negative cells provides a selective advantage to the agr negative cell population and may contribute to the common appearance of agr negative cells in S. aureus populations.

Swarm robotics has experienced a rapid expansion in recent years, primarily fueled by specialized multi-robot systems developed to achieve dedicated collective actions. These specialized platforms are, in general, designed with swarming considerations at the front and center. Key hardware and software elements required for swarming are often deeply embedded and integrated with the particular system. However, given the noticeable increase in the number of low-cost mobile robots readily available, practitioners and hobbyists may start considering to assemble full-fledged swarms by minimally retrofitting such mobile platforms with a swarm-enabling technology. Here, we report one possible embodiment of such a technology—an integrated combination of hardware and soft- ware—designed to enable the assembly and the study of swarming in a range of gener - al-purpose robotic systems. This is achieved by combining a modular and transferable software toolbox with a hardware suite composed of a collection of low-cost and off-the- shelf components. The developed technology can be ported to a relatively vast range of robotic platforms—such as land and surface vehicles—with minimal changes and high levels of scalability. This swarm-enabling technology has successfully been implemented on two distinct distributed multi-robot systems, a swarm of mobile marine buoys and a team of commercial terrestrial robots. We have tested the effectiveness of both of these distributed robotic systems in performing collective exploration and search scenarios, as well as other classical cooperative behaviors. Experimental results on different swarm behaviors are reported for the two platforms in uncontrolled environments and without any supporting infrastructure. The design of the associated software library allows for a seamless switch to other cooperative behaviors—e.g., leader–follower heading consen- sus and collision avoidance, and also offers the possibility to simulate newly designed collectiv, Singapore. National Research Foundation (Campus for Research Excellence and Technological Enterprise programme), Singapore University of Technology and Design (Seed grant #IGDS S15 01021), Singapore. Ministry of Education (grant #SUTDT12015003)

An electric vehicle’s Distance to Empty (DTE) is defined as the actual distance the vehicle can be driven before recharging is required. A real-time estimate is commonly displayed on the vehicle’s electronic instrument cluster and is used by the driver to plan their route. It is proved in this paper that the challenge for any DTE estimation algorithm is to accurately predict the future energy consumption of a vehicle. Future energy can be predicted reliably if either (i) future energy consumption is sufficiently similar to the past or (ii) applicable information about the future is known beforehand. A stochastic simulation was used to show that the average energy use (Wh/km) measured over the past ∼300 km often does an adequate job at predicting future energy use. A conventional DTE algorithm assumes this condition by “blending” both a long- and short-term average of past energy use. However, significant changes in driving conditions (e.g. traffic or auxiliary energy use) for sustained periods of time can cause large errors in DTE estimates. This paper showed that DTE error can be reduced if those future changes are detected beforehand by obtaining route information from the driver. For example, if the driver provides their destination(s) beforehand, a navigation system could obtain route, traffic and weather information via the internet and use this information to improve the DTE estimate. A multivariate linear regression model was derived that adjusts a historical average of energy consumption based on estimated changes in speeds, traffic and temperature. This method utilizes the measuring ability of the vehicle and thus does not require complex physics-based models. The algorithm could be implemented as a cloud-based mobile phone application since it is computationally light and the model is fitted using historical driving data. Finally, the algorithms were compared using a stochastic vehicle simulation and it was shown that incorporating future route information

We present a novel approach to developing a vehicle communication platform consisting of a low-cost, open-source hardware for moving vehicle data to a secure server, a Web Application Programming Interface (API) for the provision of third-party services, and an intuitive user dashboard for access control and service distribution. The CloudThink infrastructure promotes the commoditization of vehicle telematics data by facilitating easier, flexible, and more secure access. It enables drivers to confidently share their vehicle information across multiple applications to improve the transportation experience for all stakeholders, as well as to potentially monetize their data. The foundations for an application ecosystem have been developed which, taken together with the fair value for driving data and low barriers to entry, will drive adoption of CloudThink as the standard method for projecting physical vehicles into the cloud. The application space initially consists of a few fundamental and important applications (vehicle tethering and remote diagnostics, road-safety monitoring, and fuel economy analysis) but as CloudThink begins to gain widespread adoption, the multiplexing of applications on the same data structure and set will accelerate its adoption. Keyword: cloud theory; eSafety; information system; intelligent transport system; data security

Small colony variants (SCVs) of the human pathogen Staphylococcus aureus are associated with persistent infections. Phenotypically, SCVs are characterized by slow growth and they can arise upon interruption of the electron transport chain that consequently reduce membrane potential and thereby limit uptake of aminoglycosides (e.g., gentamicin). In this study, we have examined the pathways by which the fitness cost of SCVs can be ameliorated. Five gentamicin resistant SCVs derived from S. aureus JE2 were independently selected on agar plates supplemented with gentamicin. The SCVs carried mutations in the menaquinone and hemin biosynthesis pathways, which caused a significant reduction in exponential growth rates relative to wild type (WT; 0.59-0.72) and reduced membrane potentials. Fifty independent lineages of the low-fitness, resistant mutants were serially passaged for up to 500 generations with or without sub-lethal concentrations of gentamicin. Amelioration of the fitness cost followed three evolutionary trajectories and was dependent on the initial mutation type (point mutation vs. deletion) and the passage condition (absence or presence of gentamicin). For SCVs evolved in the absence of gentamicin, 12 out of 15 lineages derived from SCVs with point mutations acquired intra-codonic suppressor mutations restoring membrane potential, growth rate, gentamicin susceptibility and colony size to WT levels. For the SCVs carrying deletions, all lineages enhanced fitness independent of membrane potential restoration without alterations in gentamicin resistance levels. By whole genome sequencing, we identified compensatory mutations in genes related to the σ(B) stress response (7 out of 10 lineages). Inactivation of rpoF that encode for the alternative sigma factor SigB (σ(B)) partially restored fitness of SCVs. For all lineages passaged in the presence of gentamicin, fitness compensation via membrane potential restoration was suppressed, however, selected for secondar

BACKGROUND: In Staphylococcus aureus sub-populations of slow-growing cells forming small colony variants (SCVs) are associated with persistent and recurrent infections that are difficult to eradicate with antibiotic therapies. In SCVs that are resistant towards aminoglycosides, mutations have been identified in genes encoding components of the respiratory chain. Given the high frequencies of SCVs isolated clinically it is vital to understand the conditions that promote or select for SCVs.RESULTS: In this study we have examined how exposure to sub-inhibitory concentrations of antibiotics with different mechanism of action influence the formation of SCVs that are resistant to otherwise lethal concentrations of the aminoglycoside, gentamicin. We found that exposure of S. aureus to fluoroquinolones and mitomycin C increased the frequency of gentamicin resistant SCVs, while other antibiotic classes failed to do so. The higher proportion of SCVs in cultures exposed to fluoroquinolones and mitomycin C compared to un-exposed cultures correlate with an increased mutation rate monitored by rifampicin resistance and followed induction of the SOS DNA damage response.CONCLUSION: Our observations suggest that environmental stimuli, including antimicrobials that reduce replication fidelity, increase the formation of SCVs through activation of the SOS response and thereby potentially promote persistent infections that are difficult to treat.

Staphylococcus aureus is a human commensal that at times turns into a serious bacterial pathogen causing life-threatening infections. For the delicate control of virulence, S. aureus employs the agr quorum-sensing system that, via the intracellular effector molecule RNAIII, regulates virulence gene expression. We demonstrate that the presence of the agr locus imposes a fitness cost on S. aureus that is mediated by the expression of RNAIII. Further, we show that exposure to sublethal levels of the antibiotics ciprofloxacin, mupirocin, and rifampin, each targeting separate cellular functions, markedly increases the agr-mediated fitness cost by inducing the expression of RNAIII. Thus, the extensive use of antibiotics in hospitals may explain why agr-negative variants are frequently isolated from hospital-acquired S. aureus infections but rarely found among community-acquired S. aureus strains. Importantly, agr deficiency correlates with increased duration of and mortality due to bacteremia during antibiotic treatment and with a higher frequency of glycopeptide resistance than in agr-carrying strains. Our results provide an explanation for the frequent isolation of agr-defective strains from hospital-acquired S. aureus infections and suggest that the adaptability of S. aureus to antibiotics involves the agr locus. IMPORTANCE: Staphylococcus aureus is the most frequently isolated pathogen in intensive care units and a common cause of nosocomial infections, resulting in a high degree of morbidity and mortality. Surprisingly, a large fraction (15 to 60%) of hospital-isolated S. aureus strains are agr defective and lack the main quorum-sensing-controlled virulence regulatory system. This is a problem, as agr-defective strains are associated with a mortality level in bacteremic infections and a probability of glycopeptide resistance greater than those of other strains. We show here that agr-negative strains have a fitness advantage over agr-positive strains in the presence o