No description available.

No description available.

No description available.

No description available.

Human error has been the most common cause of car accidents. Advances in sensing and data fusion have made recent progress in autonomous vehicles that will increase the potential of drastically improving safety, efficiency, and cost of transportation. In this thesis, we present an overview of finding the error probability of sensor fusion in automotive driving, and we will investigate the collision probabilities in automated vehicles. In our study, we simulate automated driving systems in a virtual environment using real-world maps using MATLAB Automated Driving Toolbox, Simulink, and Roadrunner. During the study, we will investigate different scenarios such as weather conditions, noise, lighting, and road conditions with an ‘ego- vehicle’ equipped with multiple sensors such as; lidar and vision sensors.

One of the biggest challenges in the development ofautonomous vehicles is to anticipate the behavior of other roadusers. Autonomous vehicles rely on data obtained by on-boardsensors and make decisions accordingly, but this becomes difficultif the sensors are occluded or have limited range. In this reportwe propose an algorithm for connected vehicles in an intersectionto fuse and share sensor data and gain a better estimationof the surrounding environment. The method used for sensorfusion was a Kalman filter and a tracking algorithm, where timedelay from external sensors was considered. Parameters for theKalman filter were decided through measurement of the sensors’variances as well as tuning. It was concluded that the variancesare dependent on the objects’ movements, which means thatconstant parameters for the Kalman filter would not be enoughto make it efficient. However, the tracking and the sensor sharingmade a significant difference in the vehicle’s detection rate whichcould ultimately increase safety in intersections. / En av de största utmaningarna för utvecklingen av autonoma fordon är att förutse andra trafikanters beteenden. Autonoma fordon förlitar sig på data från sensorer ombord och fattar beslut i enlighet med informationen från dessa. Detta blir särskilt svårt om sensorerna skyms eller om sensorerna har begränsad räckvidd. I denna rapport föreslår vi en algoritm för delning och optimering av sensordata för autonoma fordon i en vägkorsning för att ge fordonet en så bra uppfattning av omgivningen som möjligt. Metoden som användes för sensorfusion var ett Kalman-filter tilsammans med en spårningsalgoritm där tidsfördröjning av data från externa sensorer togs i beaktning. Parametrarna för Kalman-filtret valdes genom mätning av sensorns varians samt genom trimning. Slutsatsen drogs att varianserna är beroende av objektens rörelsemönster, vilket innebär att konstanta parametrar för Kalman-filtret inte skulle vara tillräckligt för att göra det funktionellt. Spårningen och delningen av sensordata gjorde emellertid en betydande skillnad i andelen upptäckta objekt vilket skulle kunna nyttjas för att öka säkerheten i korsningar. / Kandidatexjobb i elektroteknik 2020, KTH, Stockholm

The aim of this thesis work is to investigate the possibility of applying a sensor fusion algorithm with a focus on estimating vehicle dynamic states, mainly the vehicle body accelerations. Modern passenger vehicles have several mechatronic systems such as active safety, comfort, driver assistance etc., which are highly dependant on accurate knowledge of such states. This work focuses on the mechatronic suspension system, which makes use of the body accelerations measurements to control the dynamics of the vehicle body in order to provide an improved driving experience. This work can be split up into two major parts, the first being the identification of available onboard sensors for measuring the vehicle body accelerations. Five different sensor combinations are considered and compared with each other. The next part is to develop a sensor fusion algorithm, in this case, a Kalman Filter (KF) based algorithm, which uses vehicle dynamic modelling knowledge to obtain accurate, reliable and less uncertain estimates of the states. Specifically, an Unscented Kalman Filter (UKF) and Cubature Kalman Filter (CKF) were built and compared with each other. Two different vehicle dynamic models, a vehicle planar dynamic model and a full car suspension model, were implemented to capture both the effects of road disturbances and drivingmanoeuvres on the vehicle body dynamics. Both these fusion algorithms were tested using simulation data and logged data and validated by comparing with an ideal sensing method to measure the body accelerations used currently at Volvo Car Corporation. / Syftet med detta examensarbete är att undersöka möjligheten att tillämpa en sensorfusionsalgoritm med fokus på att uppskatta fordonets dynamiska tillstånd, främst karossens acceleration. Moderna personbilar har flera mekatroniska system som aktiv säkerhet, komfort, förarassistans etc., som är mycket beroende av exakt kunskap om sådana tillstånd. Detta arbete fokuserar på det mekatroniska fjädringssystemet, som använder karossens accelerationsmätningar för att styra fordonets dynamik och för att ge en förbättrad körupplevelse. Detta arbete kan delas upp i två huvuddelar, den första är identifiering av tillgängliga inbyggda sensorer för mätning av fordonets accelerationer. Fem olika sensorkombinationer övervägs och jämförs med varandra. Nästa del är att utveckla en sensorfusionsalgoritm, i detta fall en kalmanfilter baserad algoritm, som använder kunskap om fordonets dynamik för att få exakta, pålitliga och mindre osäkra uppskattningar av tillstånden. Specifikt byggdes en UKF och CKF som jämfördes med varandra. Två olika fordonsdynamiska modeller, en plan dynamisk modell och en full hjulupphängningsmodell, implementerades för att fånga både effekterna av vägstörningar och körmanövrer på fordonets karossdynamik. Båda dessa fusionsalgoritmer testades med hjälp av simuleringsdata och loggade data och validerades genom att jämföra med en idealisk avkänningsmetod för att mäta karossaccelerationerna som används för närvarande på Volvo Car Corporation.

There have been several accidents involving autonomous vehicles on public roadways under scenarios that are normally avoidable by competent human drivers. This thesis contains a review of these accidents and their causes as a result of inadequate hazard mitigation. As a solution to this problem, a novel design pattern is proposed. This design pattern was developed from a hazard analysis using Systems Theoretic Process Analysis ( STPA ) methodologies that analyzed the circumstances common to several of these accidents. To demonstrate the effectiveness of the novel design pattern, an example system is constructed and tested in simulation against several accident scenarios similar to the ones studied. The results are then explained to demonstrate the effectiveness of the proposed design pattern. / Thesis / Master of Applied Science (MASc)

Yes / The advent of Internet of Things (IoT) has escalated the information sharing among various smart devices by many folds, irrespective of their geographical locations. Recently, applications like e-healthcare monitoring has attracted wide attention from the research community, where both the security and the effectiveness of the system are greatly imperative. However, to the best of our knowledge none of the existing literature can accomplish both these objectives (e.g., existing systems are not secure against physical attacks). This paper addresses the shortcomings in existing IoT-based healthcare system. We propose an enhanced system by introducing a Physical Unclonable Function (PUF)-based authentication scheme and a data driven fault-tolerant decision-making scheme for designing an IoT-based modern healthcare system. Analyses show that our proposed scheme is more secure and efficient than existing systems. Hence, it will be useful in designing an advanced IoT-based healthcare system. / Supported in part by Singapore Ministry of Education Academic Research Fund Tier 1 (R-263-000- D63-114). / Research Development Fund Publication Prize Award winner, July 2020.

The field of robotics has advanced to the point where robots are being developed for use in fire environments to perform firefighting tasks. These environments contain varying levels of fire and smoke, both of which obstruct robotic perception sensors. In order to effectively use robots in fire environments, the issue of perception in the presence of smoke and fire needs to be addressed. The goal of this research was to address the problem of perception, specifically rangefinding, in fire smoke environments. A series of tests were performed in fire smoke filled environments to evaluate the performance of different commercial rangefinders and cameras as well as a long-wavelength infrared (LWIR) stereo vision system developed in this research. The smoke was varied from dense, low temperature smoke to light, high temperature smoke for evaluation in a range of conditions. Through small-scale experiments on eleven different sensors, radar and LWIR cameras outperformed other perception sensors within both smoke environments. A LWIR stereo vision system was developed for rangefinding and compared to radar, LIDAR, and visual stereo vision in large-scale testing, demonstrating the ability of LWIR stereo vision to rangefind in dense smoke when LIDAR and visual stereo vision fail. LWIR stereo vision was further developed for improved rangefinding in fire environments. Intensity misalignment between cameras and stereo image filtering were addressed quantitatively. Tests were performed with approximately isothermal scenes and thermally diverse scenes to select subsystem methods. In addition, the effects of image filtering on feature distortion were assessed. Rangefinding improvements were quantified with comparisons to ground truth data. Improved perception in varying levels of clear and smoke conditions was developed through sensor fusion of LWIR stereo vision and a spinning LIDAR. The data were fused in a multi-resolution 3D voxel domain using evidential theory to model occupied and free space states. A heuristic method was presented to separate significantly attenuated LIDAR returns from low-attenuation returns. Sensor models were developed for both return types and LWIR stereo vision. The fusion system was tested in a range of conditions to demonstrate its ability for improved performance over individual sensor use in fire environments. / Ph. D.