FACTS - Founded in July 2008 - Meanwhile about to 250 employees. The team structure is characterized by a big number of very experienced engineers - AVL Software and Functions GmbH creates prototyping and serial solutions (software and hardware) for different applications in the fields of for example passenger cars, racing, two wheelers, light and heavy duty vehicles - AVL Software and Functions is the globally responsible competence center for software development inside theAVL group. - 100% integrated into the worldwide AVL network

Nowadays, autonomous driving revolution is getting closer to reality. To achieve the Autonomous driving the first step is to develop the Advanced Driver Assistance System (ADAS). Driver-assistance systems are one of the fastest-growing segments in automotive electronics since already there are many forms of ADAS available. To investigate state of art of development of ADAS towards Autonomous Driving, we develop Systematic Mapping Study (SMS). SMS methodology is used to collect, classify, and analyze the relevant publications. A classification is introduced based on the developments carried out in ADAS towards Autonomous driving. According to SMS methodology, we identified 894 relevant publications about ADAS and its developmental journey toward Autonomous Driving completed from 2012 to 2016. We classify the area of our research under three classifications: technical classifications, research types and research contributions. The related publications are classified under thirty-three technical classifications. This thesis sheds light on a better understanding of the achievements and shortcomings in this area. By evaluating collected results, we answer our seven research questions. The result specifies that most of the publications belong to the Models and Solution Proposal from the research type and contribution. The least number of the publications belong to the Automated…Autonomous driving from the technical classification which indicated the lack of publications in this area.

FACTS - Founded in July 2008 - Meanwhile about to 250 employees. The team structure is characterized by a big number of very experienced engineers - AVL Software and Functions GmbH creates prototyping and serial solutions (software and hardware) for different applications in the fields of for example passenger cars, racing, two wheelers, light and heavy duty vehicles - AVL Software and Functions is the globally responsible competence center for software development inside theAVL group. - 100% integrated into the worldwide AVL network

Objective: Particular testing by functional decomposition of the automated driving function can potentially contribute to reducing the effort of validating highly automated driving functions. In this study, the required size of test suites for scenario-based testing and the potential to reduce it by functional decomposition are quantified for the first time. Methods: The required size of test suites for scenario-based approval of a so-called Autobahn-Chauffeur (SAE Level 3) is analyzed for an exemplary set of scenarios. Based on studies of data from failure analyses in other domains, the possible range for the required test coverage is narrowed down and suitable discretization steps, as well as ranges for the influence parameters, are assumed. Based on those assumptions, the size of the test suites for testing the complete system is quantified. The effects that lead to a reduction in the parameter space for particular testing of the decomposed driving function are analyzed and the potential to reduce the validation effort is estimated by comparing the resulting test suite sizes for both methods. Results: The combination of all effects leads to a reduction in the test suites’ size by a factor between 20 and 130, depending on the required test coverage. This means that the size of the required test suite can be reduced by 95–99% by particular testing compared to scenario-based testing of the complete system. Conclusions: The reduction potential is a valuable contribution to overcome the parameter space explosion during the validation of highly automated driving. However, this study is based on assumptions and only a small set of exemplary scenarios. Thus, the findings have to be validated in further studies.

We investigated after effects of automation in take-over scenarios in a high-end moving-base driving simulator. Drivers performed evasive manoeuvres encountering a blocked lane in highway driving. We compared the performance of drivers 1) during manual driving, 2) after automated driving with eyes on the road while performing the cognitively demanding n-back task, and 3) after automated driving with eyes off the road performing the visually demanding SuRT task. Both minimum time to collision (TTC) and minimum clearance towards the obstacle disclosed a substantial number of near miss events and are regarded as valuable surrogate safety metrics in evasive manoeuvres. TTC proved highly sensitive to the applied definition of colliding paths, and we prefer robust solutions using lane position while disregarding heading. The extended time to collision (ETTC) which takes into account acceleration was close to the more robust conventional TTC. In line with other publications, the initial steering or braking intervention was delayed after using automation compared to manual driving. This resulted in lower TTC values and stronger steering and braking actions. Using automation, effects of cognitive distraction were similar to visual distraction for the intervention time with effects on the surrogate safety metric TTC being larger with visual distraction. However the precision of the evasive manoeuvres was hardly affected with a similar clearance towards the obstacle, similar overshoots and similar excursions to the hard shoulder. Further research is needed to validate and complement the current simulator based results with human behaviour in real world driving conditions. Experiments with real vehicles can disclose possible systematic differences in behaviour, and naturalistic data can serve to validate surrogate safety measures like TTC and obstacle clearance in evasive manoeuvres.

No description available.

A large-scale introduction of driverless trucks could start taking place during the next decade. While this could bring several economic benefits for freight transport actors and society, it may also change the freight transport system and exacerbate the negative effects of road transport. This thesis aims to increase the understanding of how an introduction of driverless trucks could materialize and impact the freight transport system in Sweden. Two overarching issues are addressed. The first is how freight transport patterns will change due to the impacts of driverless trucks on road transport supply. This is addressed in Paper 1 and Paper 2. The second issue, which is studied in Paper 3, is what factors are shaping the ongoing development towards an introduction of driverless trucks in Sweden. In Paper 1, the impact of driverless trucks on the costs for long-distance road freight transport is studied through a total cost of ownership analysis which shows that driverless trucks could enable cost reductions of around 30%-40% per ton-kilometer. A key determinant of the cost reduction is to what extent reduced driver costs will be offset by other forms of human labor that may be required for driverless truck operations. Other factors, including changes to the truck acquisition cost, have marginal importance. The cost-saving potential provides a strong motivation for freight transport actors to develop and adopt driverless trucks. In Paper 2, the impacts of driverless trucks on road transport demand, utilization of different truck types, modal split, and total logistics costs are studied by using the Swedish national freight transport model Samgods. Two scenario types are studied, one in which driverless trucks substitute manually driven trucks and one where driverless trucks capable of operating between logistics hubs are introduced as a complement to manually driven trucks. The analysis shows that in both scenarios, driverless trucks could reduce total costs for Swedish freight transport in the range of billions of SEK per year. Road transport demand and truck traffic volumes may increase significantly through modal shifts from rail and sea. This could lead to increased societal costs through, for instance, increased CO2 emissions and congestion which are, however, not quantified in the study. In Paper 3, an analysis of the innovation system of driverless trucks based on an interview study with actors involved in the development and introduction of driverless trucks in Sweden is presented. The findings suggest that there are several favorable factors for a successful introduction of driverless trucks, but also that the innovation system is characterized by a high degree of uncertainty related to what infrastructure will be required and available, what business models will be emerging, and which actors will be able to capitalize on the development and which actors that become marginalized in a future with driverless trucks. The findings from this thesis can be of interest for policymakers since it highlights potential benefits and challenges associated with driverless trucks from a transport-system perspective and the provided indicative quantitative estimates on system-level impacts offer a glimpse into a future freight transport system with driverless trucks. Also, the thesis highlights critical challenges for the innovation system of driverless trucks which could guide efforts to improve its performance. / Ett storskaligt införande av förarlösa lastbilar kan komma att inledas under det kommande årtiondet. Detta skulle kunna medföra flera nyttor för transportköpare, transportbolag och samhället i stort men kan också leda till betydande förändringar av godstransportsystemet och ökade negativa effekter från vägtransporter. Syftet med denna avhandling är att öka förståelsen för hur ett införande av förarlösa lastbilar kan ske samt påverka godstransportsystemet i Sverige. Två övergripande frågeställningar studeras. Den första är hur förarlösa lastbilar påverkar utbudet för lastbilstransporter och därigenom förändrar godstransportsystemet. Detta studeras i Artikel 1 och Artikel 2. Den andra frågeställningen är vilka faktorer som påverkar den pågående utvecklingen mot ett införande av förarlösa lastbilar, vilket studeras i Artikel 3. I Artikel 1 görs en analys av hur förarlösa lastbilar kan påverka kostnaden för långväga lastbilstransporter. Denna visar att förarlösa lastbilar kan minska den totala ägandekostnaden med runt 30-40% per tonkilometer jämfört med konventionella lastbilar. Avgörande för hur stor kostnadsbesparingen blir är i vilken utsträckning minskningar i förarkostnader vägs upp av andra lönekostnader som uppstår vid användning av förarlösa lastbilar. Andra faktorer, inklusive förändringar av inköpspriset på lastbilar, har endast marginell påverkan. Den potentiella kostnadsbesparingen utgör ett tydligt motiv för godstransportaktörer att införa förarlösa lastbilar. I Artikel 2 studeras effekterna av förarlösa lastbilar på efterfrågan på lastbilstransporter, användningen av olika lastbilstyper, fördelningen mellan transportslag, och totala transportkostnader. Analysen görs med den svenska nationella godstransportmodellen Samgods och studerar två scenariotyper. I det första scenariot ersätter förarlösa lastbilar hela flottan av konventionella lastbilar. I det andra scenariot införs förarlösa lastbilar som enbart kan köra mellan logistikterminaler som ett komplement till konventionella lastbilar. Analysen visar att förarlösa lastbilar leder till en betydande ökning av efterfrågan på lastbilstransporter till följd av överflyttningfrån sjöfart och järnväg i båda scenarierna. På nationell systemnivå kan förarlösa lastbilar minska de totala kostnaderna för svenska godstransporter i storleksordningen miljarder kronor per år. Åandra sidan kan den betydande ökningen av lastbilstrafik också medföra ökade samhällsekonomiska kostnader, till exempel genom ökade koldioxidutsläpp och trängsel, vilka dock inte kvantifieras i studien. I Artikel 3 presenteras en analys av innovationssystemet för förarlösa lastbilar som bygger på en intervjustudie med aktörer involverade i utvecklingen och införandet av förarlösa lastbilar i Sverige. Resultaten indikerar att det finns flertalet gynnsamma faktorer för ett framgångsrikt införande, samtidigt som innovationssystemet i flera avseenden karakteriseras av en låg mognadsgrad och stora osäkerheter kopplade till infrastrukturfrågor, vilka affärsmodeller som kommer uppstå samt vilka aktörer som kommer gynnas eller missgynnas av utvecklingen. Resultaten från denna avhandling kan vara av intresse för beslutsfattare då de belyser potentiella nyttor och utmaningar med förarlösa lastbilar från ett transportsystemperspektiv och de indikativa systemeffekter som kvantifieras ger en fingervisning om hur ett framtida godstransportsystem med förarlösa lastbilar kan se ut. Avhandlingen belyser också viktiga utmaningar för innovationssystemet för förarlösa lastbilar vilket kan vägleda eventuella ansträngningar för att förbättra det.

The latest trends in autonomous vehicles research gave rise to the needs for specific tools to validate and test such systems. The estimations state that to consider an autonomous vehicle statistically safe, it should drive for thousands of kilometres using traditional validation methods. This process would take a long time. Furthermore, an update in the software, would require to re-run those kilometres. Therefore, the testing must be performed exploiting virtual simulations that should realistically reflect the real world. One way to perfor msuch simulations is to let the vehicle model drive down a road map and control the surrounding traffic. To be effective, spawned traffic should not be generated too far from the target vehicle. The OpenSCENARIO standard offers a feature restricting such traffic within an ellipse centred in the central object (target vehicle). This thesis investigated what technique was more efficient and scalable to detect viable roads within the ellipse to spawn stochastic traffic on. The explored solutions are two: an analytical approach and an adaptation of the AABB tree algorithm. The research started with simple cases and incremented the scenario’s complexity during the development. Through this methodology, each technique’s positive aspects and limits have been highlighted, allowing a comparison to be made. / De senaste trenderna i autonoma fordon har ökat behovet av specifika verktyg för att validera och testa sådana system. För att kunna betrakta ett autonomt fordon som statistiskt säkert, ska enligt uppskattningar autonoma fordon köra tusentals kilometer med traditionella valideringsmetoder. Denna process skulle ta mycket lång tid. Dessutom skulle en uppdatering i mjukvaran kräva att alla dessa tusentals kilometer att körs igen. Därför måste testningen utföras med hjälp av virtuella simuleringar som bör efterlikna den reella världen realistiskt. Ett sätt att genomföra dessa simuleringar är att låta en autonom fordonsmodell köra genom ett vägnät och kontrollera kringliggande trafik. För att vara effektiv, bör kringliggande trafik inte genereras för långt bort från autonoma fordonsmodellen. OpenSCENARIO-standarden innehåller en funktion som begränsar genererad trafik inom en ellips centrerad kring fordonsmodellen. Detta examensarbete undersökte vilka tekniker som är mest effektiva och skalbara för att detektera relevanta vägar inom ellipsen att generera stokastisk trafik på. De två lösningar som studerades var: en analytisk och en numerisk som använde sig av AABB-träd-algoritmen. Utförandet började med simpla fall som successivt ökade till mer avancerade scenarion. Genom denna metodik blev varje tekniks positiva aspekter samt begränsningar belysta och jämförbara.

If automated vehicles are to be introduced on public roads, they need to be able to communicate appropriately with other road users. This can be done using various interfaces and by communicating various messages. Previous research has mainly investigated design of such communication for automated passenger cars. It is, however, currently largely unknown how corresponding communi- cation should be designed for heavy automated vehicles. Scania and RISE are collaborating in a research project on what signals need to be displayed for heavy automated vehicles when they get introduced to public roads. This thesis focuses on design of an external human-machine interface (eHMI) that conveys that a truck is operated in automated driving mode. It explored various types of message contents (abstract lights, text, symbols) as well as the effect of placement of eHMI (grille, under windshield, above windshield) and distance on understanding of the message. The emphasis was on the communi- cation to pedestrians in a crossing scenario. The thesis work was split into three design iterations according to the ”De- sign Thinking” methodology. The first iteration investigated the most preferred content types. The second investigated the effect that the screen placement on the truck had on the comprehensibility of the sign with regards to distance. The third and last iteration meant creating physical prototypes of low fidelity corresponding to the concepts from the second iteration, installing them on a truck and then evaluating them on a test track. The final evaluation was of an exploitative character and involved experts in the field of HMI design. The results showed that it was hard to interpret signals based on colors and abstract lights only. Symbols were also hard to interpret and should only be used when there is a standard for symbols for automated vehicles. Using text, on the other hand, made the message more clear and was easier to understand independently of the distance. As for the placement of the eHMI, the results show that there are preferences to using the middle and upper part of the truck.

Object detection is one of the key tasks of environment perception for highly automated vehicles. To achieve a high level of performance and fault tolerance, automated vehicles are equipped with an array of different sensors to observe their environment. Perception systems for automated vehicles usually rely on Bayesian fusion methods to combine information from different sensors late in the perception pipeline in a highly abstract, low-dimensional representation. Newer research on deep learning object detection proposes fusion of information in higher-dimensional space directly in the convolutional neural networks to significantly increase performance. However, the resulting deep learning architectures violate key non-functional requirements of a real-world safety-critical perception system for a series-production vehicle, notably modularity, fault tolerance and traceability. This dissertation presents a modular multimodal perception architecture for detecting objects using camera, lidar and radar data that is entirely based on deep learning and that was designed to respect above requirements. The presented method is applicable to any region-based, two-stage object detection architecture (such as Faster R-CNN by Ren et al.). Information is fused in the high-dimensional feature space of a convolutional neural network. The feature map of a convolutional neural network is shown to be a suitable representation in which to fuse multimodal sensor data and to be a suitable interface to combine different parts of object detection networks in a modular fashion. The implementation centers around a novel neural network architecture that learns a transformation of feature maps from one sensor modality and input space to another and can thereby map feature representations into a common feature space. It is shown how transformed feature maps from different sensors can be fused in this common feature space to increase object detection performance by up to 10% compared to the unimodal baseline networks. Feature extraction front ends of the architecture are interchangeable and different sensor modalities can be integrated with little additional training effort. Variants of the presented method are able to predict object distance from monocular camera images and detect objects from radar data. Results are verified using a large labeled, multimodal automotive dataset created during the course of this dissertation. The processing pipeline and methodology for creating this dataset along with detailed statistics are presented as well.