One central challenge for Autonomous Driving (AD) systems is ensuring functional safety. This is affected by all parts of vehicle automation systems: environment perception, decision making, and actuation. The AD system manages its activity towards achieving its goals to maintain in the safety domain, upon an environment using observation through sensors and consequent actuators. Therefore, this research investigates the operational safety for the AD system. In this research, a simulation for the Autonomous Emergency Braking (AEB) system and a simple scenario are constructed on CARLA, an open-source simulator for autonomous driving systems, to investigate the factors that impact the performance of the AEB system. The time-series data that influence the AEB are collected and fed into three time-series analysis algorithms, Autoregressive Integrated Moving Average model (ARIMA), regression tree and Long short-term memory (LSTM), to select a suitable time-series algorithm to be used for the AEB system. The results show that weather, the measurement range of the sensors, and noise can affect the results of the AEB system. After comparing the performance of these three time-series algorithms through contrasting the recall and precision of these three algorithms to detect noise in the data, the results can be obtained that LSTM has the better performance for long-term analysis. And ARIMA is more suitable for short-term time-series analysis. LSTM is chosen to analyze the time-series data, since the long-term time-series analysis is necessary for the AEB system and it can detect the noise in the variables of the AEB system with better performance. / En central utmaning för AD system är att säkerställa funktionell säkerhet. Detta påverkas av alla delar av fordonsautomatiseringssystem: miljöuppfattning, beslutsfattande och aktivering. AD -systemet hanterar sin aktivitet för att uppnå sina mål att upprätthålla inom säkerhetsområdet, i en miljö som använder observation genom sensorer och därav följande ställdon. Därför undersöker denna forskning den operativa säkerheten för AD systemet. I denna forskning konstrueras en simulering för AEB -systemet och ett enkelt scenario på CARLA, en simulator med öppen källkod för autonoma körsystem, för att undersöka de faktorer som påverkar prestandan för AEB systemet. Tidsseriedata som påverkar AEB samlas in och matas in i tre tidsserieanalysalgoritmer, ARIMA, regressionsträd och LSTM, för att välja en lämplig tidsserie-algoritm som ska används för AEB systemet. Resultaten visar att väder, mätområdet för sensorerna och brus kan påverka resultaten av AEB systemet. Efter att ha jämfört prestandan för dessa tre tidsserie-algoritmer genom att kontrastera återkallelsen och precisionen för dessa tre algoritmer för att detektera brus i data kan resultaten erhållas att LSTM har bättre prestanda för långsiktig analys. Och ARIMA är mer lämpad för korttidsanalyser i tidsserier. LSTM väljs för att analysera tidsseriedata, eftersom långsiktig tidsserieanalys är nödvändig för AEB systemet och det kan detektera bruset i variablerna i AEB system med bättre prestanda.

This dissertation examines Carla Emery's The Encyclopedia of Country Living and her contribution to the modern homesteading movement. Emery (1939-2005) advocated the use of traditional methods and antique cookbooks for food self-sufficiency. She first self published and distributed the book in "issues" (1971), which were eventually combined and sold as a single book, originally named The Old Fashioned Recipe Book (1974). A reading community developed around The Encyclopedia and readers participated in its development by submitting recipes and homesteading resources. Several editions have followed and the book remains in print, considered an authoritative reference for modern homesteading. Modern homesteading is rooted in the traditions of homesteading and domestic advice. It represents a consciousness of human relationships with the land for food production and the role domestic advice in cooking and other aspects of home economics, Emery recommended a country living foodways and 19th century methods of food production for homesteading, attempting to de-industrialize the kitchen. Emery represents an important and persistent cultural interest in going back-to-the-land. Modern homesteading practices also provide possible responses to contemporary post-industrial concerns including consumer culture, food insecurity, and the environment.

Submitted by Rafaela Souza Maldonado null (rafaela_maldonado@hotmail.com) on 2017-01-20T12:08:51Z No. of bitstreams: 1 Dissertação Rafaela M.pdf: 1654669 bytes, checksum: d62fcd1665bdff8cab105fe84c54a8e9 (MD5) / Approved for entry into archive by Juliano Benedito Ferreira (julianoferreira@reitoria.unesp.br) on 2017-01-24T16:42:41Z (GMT) No. of bitstreams: 1 maldonado_rs_me_assis.pdf: 1654669 bytes, checksum: d62fcd1665bdff8cab105fe84c54a8e9 (MD5) / Made available in DSpace on 2017-01-24T16:42:41Z (GMT). No. of bitstreams: 1 maldonado_rs_me_assis.pdf: 1654669 bytes, checksum: d62fcd1665bdff8cab105fe84c54a8e9 (MD5) Previous issue date: 2016-11-28 / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / Nos últimos anos, publicações e reedições de obras em que o testemunho é a principal característica narrativa afirmam as identidades de determinados grupos. Assim, buscamos na literatura italiana, com temática da Resistência, obras que sustentam esta ideia e uma interpretação do modo de lidar com a memória de um período significativo no contexto da Segunda Guerra. Portanto, este trabalho tem como objetivo analisar e comparar duas obras de autoria feminina nas quais as autoras reconstroem a memória partigiana a partir de suas experiências, neste episódio que foi marcante para a tradição e cultura italiana. Para isso nos embasaremos nas teorias da literatura de teor testemunhal de períodos autoritários, observando os aspectos literários; da micro-história, ressaltando as obras como materiais úteis para o estudo da História; e memorialístico, discutindo o valor da memória para a história e literatura, preservando a cultura italiana e demonstrando a característica híbrida e fronteiriça das obras. Com a apresentação destas teorias analisaremos as obras autobiográfica e diarística de Carla Capponi e Ada Gobetti, respectivamente, Con cuore di donna e Diario Partigiano. / In the last years, publications and re-editions of literary works in which the testimony is the main narrative feature affirm the identities of particular groups. Thus, we search in the Italian literature, with the theme of the Resistance, works that support this idea and an interpretation of the way of how to deal with the memory of a significant period in the World War II context. Therefore, this study aims to analyze and compare two works of female authorship in which the authors reconstruct the partisan memory from their experiences, in this remarkable episode for the Italian tradition and culture. To do so, we will rely on theories of the literature’s testimonial wording of authoritarian periods, observing the literary aspects; of the micro - history, emphasizing the works as useful materials for the study of History; and memorialistic, discussing the value of memory for the History and Literature, preserving the Italian culture and demonstrating the hybrid and frontier characteristic of the works. With the presentation of these theories we will analyze the autobiographical and diaristic works of Carla Capponi and Ada Gobetti, respectively, Con cuore di donna and Diario Partigiano. / CNPq: 168643/2014-6

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2016-06-14T15:03:32Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Ana Cecilia Gomes Mayer.docx certo.pdf: 2724679 bytes, checksum: f697f3531b4e56af6e7e351d28cbbfd0 (MD5) / Made available in DSpace on 2016-06-14T15:03:32Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Ana Cecilia Gomes Mayer.docx certo.pdf: 2724679 bytes, checksum: f697f3531b4e56af6e7e351d28cbbfd0 (MD5) Previous issue date: 2011-02-24 / CAPEs / Na natureza, cadáveres são colonizados por espécies de Diptera e Coleoptera, que se sucedem ao longo de cada estágio de decomposição. Considerando-se a escassez de inventários sobre insetos necrófagos na caatinga, esta pesquisa buscou estudar a fauna de Coleoptera associada à carcaça de suínos em área de caatinga hiperxerófila de Serra Talhada - PE. Carcaças de porcos foram colocadas em gaiolas circundadas por armadilhas de queda. Para comparação, armadilhas foram montadas em local controle. Os coleópteros associados à carcaça foram ainda coletados com pinças e por bandejas colocadas sob o animal. O experimento foi conduzido no primeiro semestre/2010. Coletaram-se 4.279 coleópteros de 13 famílias e 56 espécies, de diversos hábitos alimentares. Scarabaeidae foi a família mais abundante com 2.601 indivíduos. Deltochilum verruciferum foi a espécie mais abundante (2.407 exemplares) e dominante durante o experimento. A quantidade de insetos foi significativamente maior na carcaça que no controle (p < 0.0001). Houve diferenças significativas na riqueza e abundância ao longo da decomposição; a fase de putrefação escura atraiu maior diversidade e a fase seca, a maior abundância. Necrobia rufipes e Dermestes maculatus (necrófagos) estiveram presentes em maior abundância nas fases avançadas da decomposição. A adição de uma carcaça provoca aumento na população local de coleópteros e afeta a estrutura de comunidades. A colonização começa logo após a morte e as espécies se sobrepõem durante as fases de decomposição. / In nature, corpses are colonized by species of Diptera and Coleoptera, which succeed throughout each stage of decomposition. Considering the shortage of inventories on necrophagous insects in a region of Caatinga, this research attempted to study the fauna of Coleoptera associated with pig carcasses in an extreme enviroment of Caatinga in the municipality of Serra Talhada, state of Pernambuco, Brazil. Pig carcasses were placed inside metal cages surrounded by pitfall traps. For comparison, traps were assembled in a local far from carcasses, the control treatment. Coleopterans associated with the carcasses were also collected with tweezers and trays placed under the animals. The experiment was conducted in the first semester of 2010. 4.279 beetles were collected from 13 families and 56 species of diverse feeding habits. The Scarabaeidae family was the most abundant in richness and in number of individuals, with 2.601 collected specimens. Deltochilum verruciferum was the most abundant species (2.407 individuals) and was dominant during all the experiment. The number of insects was significantly higher in the traps around the carcasses than in control traps (p < 0.0001). There were significant differences in richness and abundance along the decomposition, the dark putrefaction stage attracted the most diversity while dry stage, the most abundant. Necrobia rufipes and Dermestes maculatus (necrophagous) were present in higher abundance in the advanced stages of decomposition. The addition of a decomposing carcass causes an increase in a local population of beetles and affects the community structure. The colonization begins soon after death and the species overlap along the decomposition stages.

Autonomous vehicles perform semantic segmentation to orient themselves, but training neural networks for semantic segmentation requires large amounts of labeled data. A hand-labeled real-life dataset requires considerable effort to create, so we instead turn to virtual simulators where the segmented labels are known to generate large datasets virtually for free. This work investigates how effective synthetic datasets are in driving scenarios by collecting a dataset from a simulator and testing it against a real-life hand-labeled dataset. We show that we can get a model up and running faster by mixing synthetic and real-life data than traditional dataset collection methods and achieve close to baseline performance.

abstract: Autonomous vehicle technology has been evolving for years since the Automated Highway System Project. However, this technology has been under increased scrutiny ever since an autonomous vehicle killed Elaine Herzberg, who was crossing the street in Tempe, Arizona in March 2018. Recent tests of autonomous vehicles on public roads have faced opposition from nearby residents. Before these vehicles are widely deployed, it is imperative that the general public trusts them. For this, the vehicles must be able to identify objects in their surroundings and demonstrate the ability to follow traffic rules while making decisions with human-like moral integrity when confronted with an ethical dilemma, such as an unavoidable crash that will injure either a pedestrian or the passenger. Testing autonomous vehicles in real-world scenarios would pose a threat to people and property alike. A safe alternative is to simulate these scenarios and test to ensure that the resulting programs can work in real-world scenarios. Moreover, in order to detect a moral dilemma situation quickly, the vehicle should be able to identify objects in real-time while driving. Toward this end, this thesis investigates the use of cross-platform training for neural networks that perform visual identification of common objects in driving scenarios. Here, the object detection algorithm Faster R-CNN is used. The hypothesis is that it is possible to train a neural network model to detect objects from two different domains, simulated or physical, using transfer learning. As a proof of concept, an object detection model is trained on image datasets extracted from CARLA, a virtual driving environment, via transfer learning. After bringing the total loss factor to 0.4, the model is evaluated with an IoU metric. It is determined that the model has a precision of 100% and 75% for vehicles and traffic lights respectively. The recall is found to be 84.62% and 75% for the same. It is also shown that this model can detect the same classes of objects from other virtual environments and real-world images. Further modifications to the algorithm that may be required to improve performance are discussed as future work. / Dissertation/Thesis / Masters Thesis Mechanical Engineering 2019

Produkter som produceras måste genomgå tester och godkännande av kontrollerande organ innan de kan säljas på marknaden. Detta är ofta en tidskrävande och ekonomisk kostsam process att utveckla en ny produkt. Hardware in loop (HIL) är ett sätt att möjliggöra tester av delsystem utan att hela systemet är komplett. Genom att använda HIL under utvecklingen av en ny produkt kan man hitta mjukvarufel samt designfel tidigt i processen, detta kan spara tid och pengar för företag. I bilindustrin började HIL system i form av en körsimulator där föraren kunde få en känsla för bilen utan köra den på vägen. Metoden lämpar sig bra för tester av autonoma drönare då deras funktion behöver valideras med rigorösa tester innan kontrollerande organ tillåter försäljning av produkterna. Risker för personskador eller prototypskador under testning kan minskas då felaktigheter troligtvis hittas och korrigeras tidigare. HIL har varit en del av utvecklingsprocesser de senaste hundra åren och har visat sig varit en effektiv metod. Under detta arbete skapades en simulerad miljö med simuleringsverktyget Carla. Carla är en simulator som bygger på spelmotorn Unreal Engine, den är framtagen med målet att användas för utveckling av tekniker för självkörande bilar. Data från den simulerade miljön användes som insignal till en HIL testbänk. Testbänken kontrollerar insignaler till en kontroll algoritm och övervakar utsignaler från drönaren som kontroll algoritmen styr. Genom att kunna tillhandahålla testbänken med simulerade data är det möjligt att utsätta kontrollalgoritmen för miljöer och omgivningar som vanligtvis hade varit besvärligt och tidskrävande att skapa i verkligheten. Utifrån det kommer vi kunna utsätta drönaren för vår testmiljö och verifiera att den fungerar som designat.

No description available.

With the rapid development of autonomous driving and artificial intelligence technology, end-to-end autonomous driving technology has become a research hotspot. This thesis aims to explore the application of deep reinforcement learning in the realizing of end-to-end autonomous driving. We built a deep reinforcement learning virtual environment in the Carla simulator, and based on it, we trained a policy model to control a vehicle along a preplanned route. For the selection of the deep reinforcement learning algorithms, we have used the Proximal Policy Optimization algorithm due to its stable performance. Considering the complexity of end-to-end autonomous driving, we have also carefully designed a comprehensive reward function to train the policy model more efficiently. The model inputs for this study are of two types: firstly, real-time road information and vehicle state data obtained from the Carla simulator, and secondly, real-time images captured by the vehicle's front camera. In order to understand the influence of different input information on the training effect and model performance, we conducted a detailed comparative analysis. The test results showed that the accuracy and significance of the information has a significant impact on the learning effect of the agent, which in turn has a direct impact on the performance of the model. Through this study, we have not only confirmed the potential of deep reinforcement learning in the field of end-to-end autonomous driving, but also provided an important reference for future research and development of related technologies.

As the autonomous vehicles are getting clo-ser to commercial roll out, the challenges for the developersof the software are getting more complex. One challenge thedevelopers are facing is the interaction between humans andautonomous vehicles in traffic.Such situation requires a hugeamount of data to in order to design and proof test autonomoussystem than can handle complex interactions with humans.Such data can not be collected in real traffic situations withoutcompromising the safety of the human counterparts, thereforesimulations will be necessary. Since human driving behavior ishard to predict, these simulations need human interaction inorder to get valid data of human behaviour.The purpose of thisproject is to develop a driving interface and then evaluate theusers experience in an experiment. To do this we have designedand implemented steering,braking and acceleration on a userinterface for a simulator used in autonomous driving researchcalled Car Learning to Act (CARLA) at the Smart Mobility Lab(SML) at KTH. We have implemented two driving simulatoruser interfaces, with different levels of information feedbackto the user. To evaluate the developed user interface, a surveywas designed to measure how intuitive the driving experiencewas while also comparing it to the original setup at SML. Thesurvey showed that the driving experience was more intuitivewith the two developed user interfaces and that 60% would feelcomfortable using the new systems on a real vehicle in traffic. / Allteftersom autonoma bilar kommer närmare kommersiell lansering blir utmaningarna för utvecklarna av mjukvaran mer komplexa. En utmaning som utvecklarna står inför är interaktionen mellan autonoma bilar och människor i och utanför trafiken. Dessa situationer kommer kräva en stor mängd data för att säkerhetställa att autonoma bilar kommer kunna agera optimalt. För att inhämta sådan data utan att riskera säkerheten för alla ute i trafiken kommer simulatorer behövas. Eftersom vi inte kan förutspå mänskligt beteende kommer industrin behöva använda mänskliga förare i dessa simulatorer för att få realistiska resultat. Syftet med detta projekt är att utveckla ett förargränssnitt för människor och sedan utvärdera autenticiten av upplevelsen från ett mänskligt perspektiv. Genom att implementera olika bilmekanismer så som styrning, inbromsning, accelerationen och retardation i en simulator för autonom bil forskning, Car Learning To Act(CARLA) i Smart Mobility Lab(SML) på KTH. Vi implementerade två användargränssnitt med olika nivåer av informations återkoppling till användaren. För att utvärdera användargränssnitten utformades ett frågeformulär för att mäta hur intuitivt körupplevelsen var och samtidigt jämföra med det originella användargränssnittet i SML. Undersökningen visade att körupplevelsen var mer intuitiv med det två utvecklade användargränssnitten och att 60% skulle vara bekväma med att använda ett utav dessa system för att styra ett riktigt fordon i trafik. / Kandidatexjobb i elektroteknik 2020, KTH, Stockholm