VEHITS 2021 Abstracts

Area 1 - Connected Vehicles

Full Papers

Paper Nr:	48
Title:	A Survey on Decentralized Cooperative Maneuver Coordination for Connected and Automated Vehicles
Authors:	Daniel Maksimovski, Andreas Festag and Christian Facchi
Abstract:	V2X communications can be applied for maneuver coordination of automated vehicles, where the vehicles exchange messages to inform each other of their driving intentions and to negotiate for joint maneuvers. For motion and maneuver planning of automated vehicles, the cooperative maneuver coordination extends the perception range of the sensors, enhances the planning horizon and allows complex interactions among the vehicles. For specific scenarios, various schemes for maneuver coordination of connected automated vehicles exist. Recently, several proposals for maneuver coordination have been made that address generic instead of specific scenarios and apply different schemes for the message exchange of driving intentions and maneuver negotiation. This paper presents use cases for maneuver coordination and classifies existing generic approaches for decentralized maneuver coordination considering implicit and explicit trajectory broadcast, cost values and space-time reservation. We systematically describe the approaches, compare them and derive future research topics.
Download

Short Papers

Paper Nr:	52
Title:	Detecting Message Modification Attacks on the CAN Bus with Temporal Convolutional Networks
Authors:	Irina Chiscop, András Gazdag, Joost Bosman and Gergely Biczók
Abstract:	Multiple attacks have shown that in-vehicle networks have vulnerabilities which can be exploited. Securing the Controller Area Network (CAN) for modern vehicles has become a necessary task for car manufacturers. Some attacks inject potentially large amount of fake messages into the CAN network; however, such attacks are relatively easy to detect. In more sophisticated attacks, the original messages are modified, making the detection a more complex problem. In this paper, we present a novel machine learning based intrusion detection method for CAN networks. We focus on detecting message modification attacks, which do not change the timing patterns of communications. Our proposed temporal convolutional network-based solution can learn the normal behavior of CAN signals and differentiate them from malicious ones. The method is evaluated on multiple CAN-bus message IDs from two public datasets including different types of attacks. Performance results show that our lightweight approach compares favorably to the state-of-the-art unsupervised learning approach, achieving similar or better accuracy for a wide range of scenarios with a significantly lower false positive rate.
Download

Paper Nr:	55
Title:	Strategic Coordination of Cooperative Truck Overtaking Maneuvers
Authors:	Jan C. Mertens, Jürgen Hauenstein, Frank Diermeyer and Andreas Zimmermann
Abstract:	This paper demonstrates how a cooperative truck overtaking maneuver can be coordinated and synchronized via V2X. This is relevant because the classical truck overtaking maneuver imposes high stress on truck drivers, which can lead to work absences or accidents. We define which abstract/atomic tasks are involved in the truck overtaking maneuver and assign them to a distributed state machine. With the help of a V2X message we then synchronize this state machine and exchange all information relevant for the overtaking maneuver. The simulation of 600 overtaking scenarios demonstrates that the developed concept is adequate and that a transmission frequency of 5 Hz offers the best trade-off between channel load and maneuver quality.
Download

Paper Nr:	76
Title:	Accelerating Interference-based QoS Analysis of Vehicular Ad Hoc Networks for BSM Safety Applications: Parallel Numerical Solutions and Simulations
Authors:	Jing Zhao, Hao Zhou, YanBin Wang, HuaLin Lu, Zhijuan Li and XiaoMin Ma
Abstract:	Vehicular Ad-hoc Networks (VANETs) have been proposed and investigated for road safety applications. Many safety applications are enabled by broadcasting basic safety message (BSM) periodically. Whether the current IEEE802.11p communication system can meet the stringent quality of service (QoS) requirement for safety applications is under discussion. Many analytical and simulation models have been proposed to study the reliability of DSRC (Dedicated Short Range Communication) IEEE802.11p broadcast services. However, most analyses assume a deterministic communication range, which is unpractical. In this paper, we propose an analytical model based on signal-to-interference-plus-noise ratio (SINR) to study of QoS and capacity of VANET for BSM based safety applications. The analytical model considers the context of the more practical vehicular communication environment: BSM broadcast, asynchronous timing between hidden terminals, Nakagami channel fading, and Non-Homogeneous Poisson Process vehicle distribution. For the proposed model, the computation complexity of QoS and capacity metrics by numerical solutions is so high that the computation time is intolerable. Thus the efficient numerical way together with a parallel approach is needed to evaluate these metrics. The Monte Carlo integration and MPI (Message Passing Interface) method are applied for accelerating the computing process. The analysis of QoS metrics are validated by NS2 simulation.
Download

Paper Nr:	31
Title:	A Cooperative Platooning Controller for Connected Vehicles
Authors:	Youssef Bichiou, Hesham Rakha and Hossam M. Abdelghaffar
Abstract:	One of the key priorities of technologies is performance. In the area of transportation, performance is typically intertwined with increased mobility and reduced costs. Congestion alleviation which is a persistent challenge faced by many cities is a priority. The use of infrastructure is inherently inefficient, resulting in higher vehicle fuel consumption and pollution. This in turn burdens commuters and businesses. Therefore, solving this issue is of prime significance because of the potential benefit. Many technologies have been and are being developed. These include adaptive traffic signals and various dynamic traffic control strategies. This paper introduces a platooning controller that keeps relatively small time gaps between consecutive vehicles to increase mobility, and eventually reduce travel costs. This controller also accounts for complex dynamic and kinematic restrictions controlling vehicle motion. The controller is tested in a virtual environment on highways in downtown Los Angeles. A drop-in travel time, delay, fuel consumption was observed across the area for connected automated vehicles (CAVs) and non-connected vehicles, at various market penetration rates (MPRs). Reductions of up to 5%, 9.4%, and 8.17% in travel time, delay, and fuel consumption, respectively are observed. These observations are observed for all vehicles platooned and non-platooned.
Download

Paper Nr:	60
Title:	A Practical Evaluation Method for Misbehavior Detection in the Presence of Selfish Attackers
Authors:	Marek Wehmer and Ingmar Baumgart
Abstract:	With recent deployment activities of Vehicle-to-X systems, the need for practical misbehavior detection is growing. The academic discussion on related topics has progressed in the last years and delivered new evaluation approaches. Most research however concentrates on evaluations based on the confusion matrix of packet classifications, such as the precision-recall graph. We show that his approach has fundamental limitations and does not allow to derive valid statements about the real-world impact of a scheme. After reviewing the state of the art, we show that the physical manifestation of attacks must be considered when evaluating misbehaviour detection systems. We propose a shift of perspective towards an attacker-oriented evaluation and contribute a new metric for selfish attackers based on the physical impact of the attack. We further present a simulation framework to practically evaluate misbehavior detection systems.
Download

Paper Nr:	70
Title:	Establishing End-to-End Secure Channel for IoT Devices through an Untrusted C-ITS Network
Authors:	Simon Bouget, Shahid Raza and Martin Furuhed
Abstract:	Critical infrastructure is becoming increasingly connected, with tighter integration to the Internet of Things (IoT). Transportation systems in particular are getting smarter with increased cooperation between vehicles and the supporting infrastructure (V2X communications), and with intelligent devices introduced in the ecosystem, either tightly integrated to the vehicle (e.g. ECUs, cameras, ...) or external sensors (e.g. temperature sensor in an attached container, smart traffic light, ...). A number of communication and security protocols are being standardized for this Cooperative Intelligent Transport Systems (C-ITS). However, using the current C-ITS standards, the security of individual devices may terminate at the gateway of a vehicle, and consequently in most existing vehicles, individual systems leak sensitive data across vendors. In this paper, we propose an end-to-end security architecture between C-ITS devices and back-end servers, in which sensitive data from individual devices can be transmitted without trusting third-parties providing the communication infrastructure (e.g. proxies, vehicle gateways, routers). The proposed solution is a standard-based integrated system that exploits recent IoT security standards and ensures inter-operability between C-ITS protocols and conventional Internet protocols. We perform a formal analysis of our architecture using the Tamarin Prover and show that it guarantees the secrecy and authenticity of the communications under adversarial settings.
Download

Paper Nr:	79
Title:	Requirements for a Cybersecurity Case Approach for the Assurance of Future Connected and Automated Vehicles
Authors:	Luis-Pedro Cobos, Alastair R. Ruddle and Giedre Sabaliauskaite
Abstract:	Cybersecurity is an issue of increasing concern for emerging connected vehicles. Ensuring public trust in future connected and automated vehicles will require very high levels of confidence in their dependability, which will include cybersecurity assurance. In functional safety engineering, the safety case has become a widely used approach to describing and documenting safety assurance arguments and their supporting evidence. The use of a similar security case can also be considered in cybersecurity engineering, but there are significant differences between safety and cybersecurity. Cybersecurity impacts include, but are not limited to, possible safety issues. Furthermore, the cybersecurity threats arise from the ingenuity of human attackers, and available technology, with the result that they are constantly evolving. This paper proposes the use of an assurance case approach for cybersecurity and outlines the particular requirements that are considered to be necessary for the development of such a cybersecurity case.
Download

Paper Nr:	94
Title:	Multi-MNO Predictive-QoS for Vehicular Applications
Authors:	Prachi Mittal and Tim Leinmüller
Abstract:	There are more and more ‘connected’ vehicles on the streets, and they run increasingly more safety critical applications. To meet the connectivity requirements of these vehicles, network providers need to not only ensure the quality of service (QoS) but also to predict any upcoming changes in the QoS and inform the vehicle(s) about it. This concept is called predictive-QoS (P-QoS) and is being heavily discussed in various organizations, e.g. 3GPP, 5GAA. To allow a seamless service to vehicles, some issues such as handling multiple mobile network operators (MNOs), while roaming for example, need to be addressed. For example, if prediction about QoS is available for multiple MNOs simultaneously for a specific area, this could be beneficial for the vehicle in selecting an MNO for further operation in specific scenarios, e.g. roaming, driving through an area where the current MNO is predictive to have poor QoS. In this paper, we introduce an entity, that takes the QoS prediction about multiple MNOs and makes decision about how to manage the connectivity in a vehicle, e.g. selecting a set of MNOs for further connectivity including a preference for each, making an “MNO usage timeplan” based on the QoS comparison etc.
Download

Area 2 - Intelligent Transport Systems and Infrastructure

Full Papers

Paper Nr:	21
Title:	An Empirical Study on Low-cost, Portable Vehicle’s Weight Estimation Solution using Smartphone’s Acceleration Data for Developing Countries
Authors:	Saima Mohan and Prashant Kumar
Abstract:	One in every three trucks in developing countries are overloaded, causing damage to roads and susceptible to accidents. Conventionally, vehicle’s weight is measured at fixed weigh stations and result in high traffic congestions at toll booths. To improve highway traffic and enhance regulation, we propose a low-cost, portable sensor-based system viable for continuous real-time assessment of vehicle’s weight. A smartphone-based sensing device is installed in vehicle and weight is estimated by applying multiple linear regression model on acceleration data. In this paper, we include statistical features having relationship with target variable. A consistent model performance of vehicle’s weight estimated at all speed ranges is established; we also evaluate the improvised model under engine idling state. An increased accuracy is obtained with error of 2% in engine idling state and overall system error of 6% with vehicle in motion. A heterogenous data source (such as vehicle class, load condition, goods, sensor locations, etc.,) of vehicle operating on Indian highway segment are collected to evaluate model robustness. With exploitation of big data and advanced analytics; advent of this solution will leverage contribution in Intelligent Transport System, focused towards smart and sustainable transportation for ASEAN region.
Download

Paper Nr:	27
Title:	Turning Rate Estimation in Roundabouts: Analysis and Validation of Different Estimation Methods
Authors:	Mánuel Gressai and Tamás Tettamanti
Abstract:	The knowledge of turning rates in roundabouts is a crucial element of traffic modeling. Measuring the turning movements is often carried out by manual traffic counts (noting on paper or using handheld devices), which is a labor-intensive, therefore expensive process. The aim of this paper is the examination and comparison of different estimation methods used for turning rates in roundabouts. Traditional iteration based approach as well as estimators adopted from control theory are discussed, benchmarked, and validated on real-world traffic data. For the estimation procedures, the traffic flows (measured at each leg of the intersection) are the input. In this way, the traditional origin-destination traffic count at an intersection can be substituted by automated traffic detection at the cross-sections together with the adequately implemented estimation process (suggested in the paper). The calibration of estimation methods is of crucial importance as well. The calibration is demonstrated based on real-world traffic counts at roundabouts. The different methods have been compared using different error metrics. As a main finding of the research, it is shown that, given the right tuning, constrained Kalman Filtering outperforms the unconstrained Kalman Filtering and the traditional iterative procedure.
Download

Paper Nr:	40
Title:	Dynamic and Continuous Berth Allocation using Cuckoo Search Optimization
Authors:	Sheraz Aslam, Michalis P. Michaelides and Herodotos Herodotou
Abstract:	Over the last couple of decades, demand for seaborne containerized trade has increased significantly and it is expected to continue growing over the coming years. As an important node in the maritime industry, a maritime container terminal (MCT) should be able to tackle the growing demand for sea trade. Due to the increased number of ships that can arrive simultaneously at an MCT combined with inefficient berth allocation procedures, there are often undesirable situations when the ships have to stay in waiting queues and delay both their berthing and departure. In order to improve port efficiency in terms of reducing the total handling cost and late departures, this study investigates the dynamic and continuous berth allocation problem (DC-BAP), where vessels are assigned dynamically as they arrive at their berth locations assuming a continuous berth layout. First, the DC-BAP is formulated as a mixed-integer linear programming (MILP) model. Since BAP is an NP-hard problem and cannot be solved by mathematical approaches in a reasonable time, this study adopts the recently developed metaheuristic cuckoo search algorithm (CSA) to solve the DC-BAP. For validating the performance of the proposed CSA method, we use a benchmark case study and a genetic algorithm solution proposed in recent literature as well as compare our results against the optimal MILP solution. From the simulation results, it becomes evident that the newly proposed algorithm has higher efficiency over counterparts in terms of optimal berth allocation within reasonable computation time.
Download

Paper Nr:	56
Title:	A Reinforcement Learning Approach for Traffic Control
Authors:	Urs Baumgart and Michael Burger
Abstract:	Intelligent traffic control is a key tool to achieve and to realize resource-efficient and sustainable mobility solutions. In this contribution, we study a promising data-based control approach, reinforcement learning (RL), and its applicability to traffic flow problems in a virtual environment. We model different traffic networks using the microscopic traffic simulation software SUMO. RL-methods are used to teach controllers, so called RL agents, to guide certain vehicles or to control a traffic light system. The agents obtain real-time information from other vehicles and learn to improve the traffic flow by repetitive observation and algorithmic optimization. As controller models, we consider both simple linear models and non-linear radial basis function networks. The latter allow to include prior knowledge from the training data and a two-step training procedure leading to an efficient controller training.
Download

Paper Nr:	61
Title:	Urban Traffic Incident Detection for Organic Traffic Control: A Density-based Clustering Approach
Authors:	Ingo Thomsen, Yannick Zapfe and Sven Tomforde
Abstract:	The traffic demands in urban road networks can fluctuate immensely. The Organic Traffic Control (OTC) offers a resilient traffic management to control such traffic demands. An additional challenge is the detection of unforeseen traffic incidents. To enhance the capabilities of OTC accordingly, we outline a traffic incident algorithm based on DBSCAN, a density-based clustering algorithm: In a simulated urban road network, equipped with traffic light controllers at intersections, vehicle detectors are used to gather traffic flow data. The clustering of this time series data to detect simulated road blockages is expanded using various filters. This extension of the initial clustering is the result of an manual evaluation process, which shows the principal applicability of this approach.
Download

Paper Nr:	69
Title:	Ride-hailing Emissions Modeling and Reduction through Ride Demand Prediction
Authors:	Tanmay Bansal, Ruchika Dongre, Kassie Wang and Sam Fuchs
Abstract:	Transportation is the largest contributor of greenhouse gas emissions in the United States. As Transportation Network Companies (TNCs), such as Uber and Lyft, grow in prevalence, it is imperative to quantify their emissions impact. We studied the case of Austin, Texas through its primary ride-hailing service - RideAustin - that has released data on 1.4+ million individual rides over an 11-month period. We estimated a total of 6014.95 metric tonnes of CO2 emissions from deadheading (when there are no passengers in freight) over the given time period. We clustered Austin into different zones and built an LSTM-based neural network for hourly ride demand forecasting on each zone through spatiotemporal features (weather, federal holidays, day of the week, and a look-back interval). Despite a large out-of-time validation window (7 months), our model outperforms the XGBoost-based baseline model by 34.86% and the next best comparable model in current literature by 15.3% in terms of MAE. In addition, we estimated a 10.624% reduction in total deadheading emissions for the same period given that the ride-hailing drivers on road are routed according to the proposed hourly ride demand forecasts.
Download

Short Papers

Paper Nr:	23
Title:	Value Networks and Monetization Strategies for C-ITS Safety Use Cases
Authors:	Pol Camps-Aragó, Simon Delaere and Ruben D’Hauwers
Abstract:	C-ITS safety use cases promise to reduce road accidents. However, deploying the necessary system elements that enable such use cases entails challenges in terms of value network coordination, return on investments in infrastructure and in-vehicle devices, and monetization of services. In short, this paper aims at contributing to overcome these economic challenges by (i) clarifying the overall value network and interactions amongst key stakeholders, (ii) proposing how to incentivise the fulfilment of bottleneck value network roles, (iii) providing recommendations on how to incentivise investments and the monetization of C-ITS services, and (iv) arguing for a data exchange and governance model based on regulatory and business model aspects.
Download

Paper Nr:	39
Title:	A Self-organising System Combining Self-adaptive Traffic Control and Urban Platooning: A Concept for Autonomous Driving
Authors:	Heiko Hamann, Julian Schwarzat, Ingo Thomsen and Sven Tomforde
Abstract:	Platooning is an approach to coordinate the driving behaviour of vehicles on major roads such as motorways. The aim is to take advantage of, e.g., slipstream effects to reduce cost. We present an approach to transfer the platooning concept to urban road networks of cities. The reduced slipstream effect is compensated by integration with the signalisation infrastructure to dynamically allow for prioritisation of platoons using progressive signal systems (i.e., “green waves”). We define the scenario and derive a research road map towards fully self-organised platoon operations and integrated coordination with self-adaptive and self-organising urban traffic control systems. Starting from both directions, that is, self-organised urban platooning as well as self-organised progressive signal systems in urban road networks, we define the scenario, identify main challenges, and present first results to demonstrate the feasibility of our research agenda.
Download

Paper Nr:	72
Title:	Evaluation of Passenger Car Emission Indexes in Relation to Passing through the Rail-road Crossing
Authors:	Mateusz Nowak, Maciej Andrzejewski, Sylwin Tomaszewski, Paweł Daszkiewicz and Patryk Urbański
Abstract:	One of the crucial aspects in the vehicles exhaust emission is, often the long duration of vehicles stop phase before the rail-road crossing before the rail vehicle passes through. During the road vehicle stop, the combustion engine in most cases operates in idling conditions. Some drivers turns off the combustion engines during mentioned stop time. In modern vehicles there is also the start&stop system implemented, which automatically stops and starts the combustion engine. Combustion engine idling phase is related to inefficient operation, where after switching the engine off, the catalytic converter could cool down, which could result in increased emission of harmful exhaust compounds after start of the engine. The analysis made in reference to the Poznan agglomeration, shows many places where alternative routes can be determined with regard to the necessity of reaching the destination, when there is a road-rail crossing on the way. The purpose of the performed work was first of all to determine the potential of reducing fuel consumption and exhaust emission by cars as a result of the improvement in the transport system efficiency. The improvement of transport system efficiency could be assumed as a trip duration reduction, when the driver could receive the information about the actual state of the rail-road infrastructure.
Download

Paper Nr:	96
Title:	Towards a Natural Language Dialog System for Mobility Service Platforms
Authors:	David Thulke, Felix Schwinger and Karl-Heinz Krempels
Abstract:	Due to a rise in novel mobility modes, urban transportation systems have become more heterogeneous and complicated in recent years. Mobility Service Platforms integrate different mobility services to offer integrated travel information, booking, and travel assistance, regardless of mobility provider or mode. Traditionally, users access these information systems through graphical user interfaces. Especially for the older population, such a sophisticated information system for a complex problem is problematic. Therefore, in this paper, we propose an approach and a prototype for a natural language interface for Mobility Service Platforms. The natural language interface allows access to the Mobility Service Platforms’ information systems and integrates other domains, such as event and place information into natural language queries. To this end, we introduce a simple unified data model for travel, event, and Point of Interest domain and design an interaction model for the natural language interface. We evaluate the prototype in a case study with potential users. The evaluation shows that most users are more comfortable interacting with a mobility service platform using natural language instead of using different graphical user interfaces providing similar functionality.
Download

Paper Nr:	11
Title:	Driving Behavior Analysis and Traffic Improvement using Onboard Sensor Data and Geographic Information
Authors:	Jun-Zhi Zhang and Huei-Yung Lin
Abstract:	In this paper, we present a method to extract the training and testing data from geographic information system (GIS) and global position system (GPS) for neural networks. Traffic signs, traffic lights and road information from the OpenStreetMap (OSM) and the government platform are compared with driving data and videos to extract images containing the important information. We also propose traffic improvement suggestions for intersections or roads by analyzing the relationship between driving behaviors, traffic lights, and road infrastructures. We use OBD-II and CAN bus logger to record more driving information, such as engine speed, vehicle speed, steering wheel steering angle, etc. We analyze the driving behavior using sparse automatic encoders and data exploration to detect abnormal and aggressive behavior. The relationship between the aggressive driving behavior and road facilities is derived by regression analysis, and some suggestions are provided for improving specific intersections or roads.
Download

Paper Nr:	22
Title:	An Intelligent Transportation System for Air and Noise Pollution Management in Cities
Authors:	Mariam O. Zaky and Hassan Soubra
Abstract:	Air quality and noise levels in urban cities have become major environmental concerns worldwide. Road vehicles are a primary source of air pollution in urban cities and they also are a considerable source of noise pollution. Undeniably, air and noise pollution are hazardous to human health. Managing pollution levels has become an absolute priority in order to reduce the anthropogenic impact on the environment. In this paper, we propose an Intelligent Transportation System-ITS for monitoring and managing air and noise pollution caused by road vehicles in cities. The system proposed dynamically routes a vehicle using, on the one hand, its particle emissions and noise indicators; and on the other hand, a city’s pollution levels and defined thresholds. The system proposed in this paper could be used for pollution based road tolls or taxes.
Download

Paper Nr:	77
Title:	A Comprehensive View of Intelligent Transport Systems and Supply Chain Management for CIS Countries
Authors:	Onur Guvenc
Abstract:	The application of Intelligent Transportation Systems (ITS) has continued to revolutionize contemporary Supply Chain Management (SCM). Operators leverage real-time information about vehicle conditions and traffic to improve the overall transport systems through Global Positioning System data, among other avenues. Such systems have been invaluable in commercial vehicle administrative processes, vehicle clearance, automated roadside safety inspection, hazardous materials incident reporting, and commercial fleet administration and management. Using a sample of 500 respondents, the study used a mixed method design to understand the key challenges faced by the intelligent transport systems and supply chain management processes in the Commonwealth of Independent States countries. The study realized that concerns about security of goods, threats from external attacks, and transportation problems are key concerns among users. The study finds a statistically significant association between cases of loss of goods and the SCM applications rating, but no statistically significant association between the cases of loss of goods and the type of goods and the type of goods delivered. Infrastructural development of countries provides comparative advantages in the use of ITS systems. Future studies should attempt to use comprehensive data.
Download

Paper Nr:	78
Title:	Transit Performance Evaluation at Signalized Intersections of Bus Rapid Transit Corridors
Authors:	Robel Desta, Tewodros Dubale and János Tóth
Abstract:	Bus Rapid Transit (BRT) is one of the mass transportation solutions consisting of infrastructures integrating dedicated bus lanes and smart operational service with different ITS technologies like Transit Signal Priority (TSP). Delay at an intersection is among the major factors for poor transit performance. This study examines the performance of buses at intersections of BRT corridors, which are privileged with Signal Priority on the dedicated lane. Simulation models were developed for the selected intersection together with the real-time calibration and validation. Statistical comparisons were conducted to test the alternative scenarios aimed at visualizing the deployment advantages. TSP options were evaluated by using PTV VISSIM with VisVAP add-on simulation tool. Alternative scenarios with and without TSP were tested to measure the performance of BRT buses along with impact assessment on the general traffic. TSP reduces travel time and control delay, improves travel speed and the results depicted a reduction in average passenger delay by 10–20%. The improvement on travel speed at an intersection of BRT vehicles were determined to be 6–8%. Prioritizing buses has diminutive impact on the general traffic, nonetheless, it is the easiest way of improving transit performance.
Download

Paper Nr:	80
Title:	Collection of Requirements and Model-based Approach for Scenario Description
Authors:	Thilo Braun, Lennart Ries, Franziska Körtke, Lara Turner, Stefan Otten and Eric Sax
Abstract:	As the level of automation and variety of Advanced Driver Assistance Systems (ADAS) and Automated Driving (AD) increases, new challenges for Verification and Validation (V&V) methods emerge. This applies especially in urban areas due to the combination of many different environmental elements, participant types, and interactions between the participants. Scenario-based testing and resimulation of recorded data are promising approaches to tackle these new challenges. An elementary component of these methods is the scenario description, which serves as a connection between different working steps in the V&V workflow. This heterogeneous usage of the scenarios during the development and validation process leads to a multitude of different, sometimes contradictory, demands on the scenario description. Nevertheless, a uniform description is desirable for easy exchange and automation. The contribution of this paper is twofold: Firstly, the described versatile field of demands is systematically broken down to requirements for the scenario description languages. This step is essential to ensure broad applicability. Secondly, this paper introduces a holistic scenario description language that is usable for generation, extraction from real-world test drives and execution of the scenarios enabling an automated workflow and increased traceability between generated, extracted and resimulated scenarios. The description and uses a model based approach and has been exemplarily tested for manually created scenarios and automatic resimulation of real-word test drives.
Download

Area 3 - Intelligent Vehicle Technologies

Full Papers

Paper Nr:	1
Title:	Unconstrained License Plate Detection in Hardware
Authors:	Petr Musil, Roman Juránek and Pavel Zemčík
Abstract:	In this paper, we propose an FPGA implementation of license plate detection (LPD) in images captured by arbitrarily placed cameras, vehicle-mounted cameras, or even handheld cameras. In such images, the license plates can appear in a wide variety of positions and angles. Thus we cannot rely on a-priori known geometric properties of the license plates as many contemporary applications do. Unlike the existing solutions targeted for DSP, FPGA or similar low power devices, we do not make any assumptions about license plate size and orientation in the image. We use multiple sliding window detectors based on simple image features, each tuned to a specific range of projections. On a dataset captured by a camera mounted on a vehicle, we show that detection rate is 98 % (and 98.7 % when combined with video tracking). We demonstrate that our FPGA implementation can process 1280×1024 pixel image at over 40 FPS with a minimum width of detected license plates approximately 100 pixels. The FPGA block is fully functional and it is intended to be used in a smart camera to parking control in residential zones.
Download

Paper Nr:	13
Title:	Radar Artifact Labeling Framework (RALF): Method for Plausible Radar Detections in Datasets
Authors:	Simon T. Isele, Marcel P. Schilling, Fabian E. Klein, Sascha Saralajew and J. M. Zoellner
Abstract:	Research on localization and perception for Autonomous Driving is mainly focused on camera and LiDAR datasets, rarely on radar data. Manually labeling sparse radar point clouds is challenging. For a dataset generation, we propose the cross sensor Radar Artifact Labeling Framework (RALF). Automatically generated labels for automotive radar data help to cure radar shortcomings like artifacts for the application of artificial intelligence. RALF provides plausibility labels for radar raw detections, distinguishing between artifacts and targets. The optical evaluation backbone consists of a generalized monocular depth image estimation of surround view cameras plus LiDAR scans. Modern car sensor sets of cameras and LiDAR allow to calibrate image-based relative depth information in overlapping sensing areas. K-Nearest Neighbors matching relates the optical perception point cloud with raw radar detections. In parallel, a temporal tracking evaluation part considers the radar detections’ transient behavior. Based on the distance between matches, respecting both sensor and model uncertainties, we propose a plausibility rating of every radar detection. We validate the results by evaluating error metrics on semi-manually labeled ground truth dataset of 3.28·106 points. Besides generating plausible radar detections, the framework enables further labeled low-level radar signal datasets for applications of perception and Autonomous Driving learning tasks.
Download

Paper Nr:	42
Title:	Soft Fault Detection and Localization in an Unshielded Twisted Pair Network using Power Line Communication
Authors:	Abdel Karim Abdel Karim, Virginie Degardin, Vincent Cocquempot and M. Amine Atoui
Abstract:	Vehicular electrical wires and communication systems can be affected by faults. Depending on their severity, faults can be divided into two families: hard and soft faults. Hard faults indicate open/short circuits that can lead to serious breakdowns because they prevent the flow of energy or information. Even though soft faults do not have such severe impacts on the system, they may develop into hard faults in the long term, hence the need to detect them. In this paper, an unshielded twisted pair cable that undergoes a water-tree degradation is considered. A soft fault, which may correspond to the effect of a mechanical constraint on the cable, is introduced as a series resistor. The studied network is a tree-shape network composed of one source and multiple receivers, one at each end of a branch. Assuming that these receivers operate in a healthy state, to detect the fault, the transfer function from each receiver are monitored and a detection index is used. Another index, based on the comparison of the effect of the fault at each endpoint, is proposed to locate the affected branch. To summarize the detection and localization algorithm, a signature matrix is generated. Simulation results are presented to illustrate our approach.
Download

Paper Nr:	43
Title:	Water Hazard Depth Estimation for Safe Navigation of Intelligent Vehicles
Authors:	Zoltan Rozsa, Marcell Golarits and Tamas Sziranyi
Abstract:	This paper proposes a method to provide depth information about water hazards for ground vehicles. We can estimate underwater depth even with a moving mono camera. Besides the physical principles of refraction, the method is based on the theory of multiple-view geometry and basic point cloud processing techniques. We use the information gathered from the surroundings of the hazard to simplify underwater shape estimation. We detect water hazards, estimate its surface and calculate real depth of underwater shape based on matched points using refraction principle. Our pipeline was tested on real-life experiments, on-board cameras and a detailed evaluation of the measurements is presented in the paper.
Download

Paper Nr:	49
Title:	Point Cloud based Hierarchical Deep Odometry Estimation
Authors:	Farzan E. Nowruzi, Dhanvin Kolhatkar, Prince Kapoor and Robert Laganiere
Abstract:	Processing point clouds using deep neural networks is still a challenging task. Most existing models focus on object detection and registration with deep neural networks using point clouds. In this paper, we propose a deep model that learns to estimate odometry in driving scenarios using point cloud data. The proposed model consumes raw point clouds in order to extract frame-to-frame odometry estimation through a hierarchical model architecture. Also, a local bundle adjustment variation of this model using LSTM layers is implemented. These two approaches are comprehensively evaluated and are compared against the state-of-the-art.
Download

Paper Nr:	51
Title:	Study of Stability through Lyapunov Theory and Passivity following a FDI on a Velocity Control System
Authors:	M. Ruhnke, X. Moreau, A. B. Neto, M. Moze, F. Aioun and F. Guillemard
Abstract:	Ensuring safety and fault tolerant strategies is essential in the development of Advanced Driver Assistance System, such as an automated cruise control.This work presents a study of the stability of switched regulated systems following the reconfiguration of the speed controller due to a fault. Firstly, the context of these works is presented highlighting the need to have a fault management system with a diagnostic part and a reconfiguration part in order to ensure the operating safety. The reconfiguration part can take the form of a switch thus involving the study of stability. It is in this context that, secondly, the passivity of the plant as well as of both the controllers (CRONE and PI) is demonstrated. As the switch takes place between two elements of a passive nature, the last point of this work highlights the application of the continuous approach in order to demonstrate the passivity and therefore the stability of the regulated plant despite the presence of the switch. To address this problem, an augmented model in the form of a generic state space representation of the controllers and the plant is constructed. Then, a Lyapunov candidate function representing the sum of the storage function of the controller and the plant is defined. A sign study of this function as well as its derivative is carried out for the two operational modes (CRONE regulating the plant and PI regulating the plant) in order to demonstrate the passivity of the switched regulated systems.
Download

Paper Nr:	59
Title:	Effects on Traffic Performance Due to Heterogeneity of Automated Vehicles on Motorways: A Microscopic Simulation Study
Authors:	Ivan Postigo, Johan Olstam and Clas Rydergren
Abstract:	The introduction of automated vehicles (AVs) is commonly expected to improve different aspects of transportation. A long transition period is expected until AVs become prevalent on roads. During this period, different types of AVs with different driving logics will coexist along human driven vehicles. Using microscopic traffic simulation, this study investigates the range of potential impacts on traffic performance in terms of throughput and travel delays for different types of AVs and human driven vehicles on motorways. The simulation experiment includes scenarios with combinations of three different driving logics for AVs together with human driven vehicles at increasing penetration rates. The utilized AV driving logics represent the evolution of AVs, they were defined in the microscopic simulation tool Vissim and were created by modifying and extending the human driver behaviour models. The results of the simulation experiment show a decrease in vehicle throughput and significant effects on delay times when AVs with a more cautious driving logic are predominant. Overall, results show higher vehicle throughput and lower travel delays as AVs evolve to more advanced driving logics.
Download

Paper Nr:	65
Title:	Comparison of Camera-Equipped Drones and Infrastructure Sensors for Creating Trajectory Datasets of Road Users
Authors:	Amarin Kloeker, Robert Krajewski and Lutz Eckstein
Abstract:	Due to the complexity of automated vehicles, their development and validation require large amounts of naturalistic trajectory data of road users. In addition to the classical approach of using measurement vehicles to generate these data, approaches based on infrastructure sensors and drones have become increasingly popular. While advantages are postulated for each method, a practical comparison of the methods based on measurements of real traffic has so far been lacking. We present a theoretical and experimental analysis of two image-based measurement methods. For this purpose, we compare measurements of a drone-based system with a prototypical camera-based infrastructure sensor system. In addition to the detection statistics of the road users, the detection quality of both systems is also investigated using a reference vehicle equipped with an inertial navigation system. Through these experiments, we can confirm each approach’s advantages and disadvantages emerging from the theoretical analysis.
Download

Paper Nr:	67
Title:	B-ETS: A Trusted Blockchain-based Emissions Trading System for Vehicle-to-Vehicle Networks
Authors:	Lam D. Nguyen, Amari N. Lewis, Israel Leyva-Mayorga, Amelia Regan and Petar Popovski
Abstract:	Urban areas are negatively impacted by Carbon Dioxide (CO2) and Nitrogen Oxide (NOx) emissions. In order to achieve a cost-effective reduction of greenhouse gas emissions and to combat climate change, the European Union (EU) introduced an Emissions Trading System (ETS) where organizations can buy or receive emission allowances as needed. The current ETS is a centralized one, consisting of a set of complex rules. It is currently administered at the organizational level and is used for fixed-point sources of pollution such as factories, power plants, and refineries. However, the current ETS cannot efficiently cope with vehicle mobility, even though vehicles are one of the primary sources of CO2 and NOx emissions. In this study, we propose a new distributed Blockchain-based emissions allowance trading system called B-ETS. This system enables transparent and trustworthy data exchange as well as trading of allowances among vehicles, relying on vehicle-to-vehicle communication. In addition, we introduce an economic incentive-based mechanism that appeals to individual drivers and leads them to modify their driving behavior in order to reduce emissions. The efficiency of the proposed system is studied through extensive simulations, showing how increased vehicle connectivity can lead to reduction of the emissions generated from those vehicles. We demonstrate that our method can be used for full life-cycle monitoring and fuel economy reporting. This leads us to conjecture that the proposed system could lead to important behavioural changes among the drivers.
Download

Paper Nr:	68
Title:	A Comparison of Lateral Intention Models for Interaction-aware Motion Prediction at Highways
Authors:	Vinicius Trentin, Antonio Artuñedo, Jorge Godoy and Jorge Villagra
Abstract:	To safely navigate in complex scenarios is crucial to know the predictions of the vehicles involved in the scene. The future behavior of the traffic participants is dependent on their intentions, the road layout and the interaction between them. In this work, a framework is presented to compute the motion predictions of the surrounding vehicles considering all possible routes obtained from a given map. At each time step, with a Dynamic Bayesian Network, the probability of being on a specific route and the intention to change lanes are computed. Our framework, based on Markov chains, is generic and can handle various road layouts and any number of vehicles. We apply the framework in a two-lane highway and evaluate the influence of different lane-changing methods on the predictions of the vehicles present at the scene.
Download

Paper Nr:	81
Title:	Pixel Invisibility: Detect Object Unseen in Color Domain
Authors:	Yongxin Wang and Duminda Wijesekera
Abstract:	Deep neural networks have been very successful in image recognition. In order for those results to be useful for driving automatons require quantifiable safety guarantees during night, dusk, dawn, glare, fog, rain and snow. In order to address this problem, we developed an algorithm that predicts a pixel-level invisibility map for color images that does not require manual labeling - that computes the probability that a pixel/region contains objects that are invisible in color domain, during light challenged conditions such as day, night and fog. We do so by using a novel use of cross modality knowledge distillation from color to thermal domain using weakly-aligned image pairs obtained during daylight and construct indicators for the pixel-level invisibility by mapping both the color and thermal images into a shared space. Quantitative experiments show good performance of our pixel-level invisibility masks and also the effectiveness of distilled mid-level features on object detection in thermal imagery.
Download

Paper Nr:	83
Title:	Deep Learning Classifiers for Automated Driving: Quantifying the Trained DNN Model’s Vulnerability to Misclassification
Authors:	Himanshu Agarwal, Rafal Dorociak and Achim Rettberg
Abstract:	The perception-based tasks in automated driving depend greatly on deep neural networks (DNNs). In context of image classification, the identification of the critical pairs of the target classes that make the DNN highly vulnerable to misclassification can serve as a preliminary step before implementing the appropriate measures for improving the robustness of the DNNs or the classification functionality. In this paper, we propose that the DNN’s vulnerability to misclassifying an input image into a particular incorrect class can be quantified by evaluating the similarity learnt by the trained model between the true class and the incorrect class. We also present the criteria to rank the DNN model’s vulnerability to a particular misclassification as either low, moderate or high. To argue for the validity of our proposal, we conduct an empirical assessment on DNN-based traffic sign classification. We show that upon evaluating the DNN model, most of the images for which it yields an erroneous prediction experience the misclassifications to which its vulnerability was ranked as high. Furthermore, we also validate empirically that all those possible misclassifications to which the DNN model’s vulnerability is ranked as high are difficult to deal with or control, as compared to the other possible misclassifications.
Download

Paper Nr:	91
Title:	A Dependency-based Combinatorial Approach for Reducing Effort for Scenario-based Safety Analysis of Autonomous Vehicles
Authors:	Kaushik Madala, Hyunsook Do and Carlos Avalos-Gonzalez
Abstract:	For an autonomous vehicle, to assure safety, we need to perform a thorough analysis considering the vehicle’s intended operational design domain (ODD). This requires analysts and engineers to consider various operating environments (OEs) that can occur in the ODD, and the various scenarios that are possible within each OE. However, the automotive safety standards ISO 26262 and ISO 21448 do not offer in-depth guidance on what and how many scenarios need to be analyzed to ensure safety of a vehicle. Moreover, many existing simulation tools and verification approaches consider limited OEs and generate test cases exhaustively for each scenario created by engineers within an OE. Such an analysis requires a significant amount of time and effort, but it still cannot ensure that various dependencies among ODD elements are covered. To address these limitations, we propose a dependency-based combinatorial approach (DBCA), which uses in-parameter- order-general (IPOG), a combinatorial testing algorithm to generate OEs and test cases for each scenario. To evaluate DBCA, we applied it to the ODD elements extracted from ISO 21448, and to a highway cut-in scenario. Our results show that DBCA reduced time and effort for analysis, and reduced the the number of OEs and test cases for the scenario without missing dependencies.
Download

Paper Nr:	97
Title:	Estimation of the Acoustic Waste Energy Harvested from Diesel Single Cylinder Engine Exhaust System
Authors:	Claudiu Golgot, Nicolae Filip and Lucian Candale
Abstract:	Noise generated in the operation of an internal combustion engine is an energy waste that produces noise pollution. Recovering some of this energy and transforming it into another form of usable energy brings significant benefits. We proposed to develop a device to recover this energy waste produced by the internal combustion engines, in the gases changing process. The developed energy recovery system is based on the Helmholtz resonator principle. For the conversion of acoustic waves into electricity, we used an audio speaker as a low-cost electromagnetic transducer located at the end of the resonant chamber. By audio playback of the acoustic signal recorded at the engine exhaust, we measured the electricity generated with the proposed recovery system. We found that the noise level measured at the exhaust depending on the engine speed range, follows a linear distribution law, instead, the harvested electric power varies nonlinearly. To find out the cause of the electric power variation, we performed a detailed FFT analysis. We found that at most engine speeds, the dominant amplitudes in the frequency spectrum are close to the resonant frequency of the system. With the proposed conversion system, we obtained a maximum value of the harvested electric power of 165 µW.
Download

Short Papers

Paper Nr:	4
Title:	Vocation Identification for Heavy-duty Vehicles: A Tournament Bracket Approach
Authors:	Daniel Kobold Jr., Andy Byerly, Rishikesh M. Bagwe, Euzeli D. Santos Jr. and Zina Ben Miled
Abstract:	The identification of the vocation of an unknown heavy-duty vehicle is valuable to parts’ manufacturers. This study proposes a methodology for vocation identification that is based on clustering techniques. Two clustering algorithms are considered: K-Means and Expectation Maximization. These algorithms are used to first construct the operating profile of each vocation from a set of vehicles with known vocations. The vocation of an unknown vehicle is then determined by using one-versus-all or one-versus-one assignment. The one-versus-one assignment is more desirable because it scales with an increasing number of vocations and requires less data to be collected from the unknown vehicles. These characteristics are important to parts’ manufacturers since their parts may be installed in different vocations. Specifically, this paper compares the one-versus-one bracket and the one-versus-one round-robin tournament assignments to the one-versus-all assignment. The tournament assignments are able to scale with an increasing number of vocations. However, the bracket assignment also benefits from a linear time complexity. The results show that despite its scalability and computational efficiency, the bracket vocation identification model has a high accuracy and a comparable precision and recall. The NREL Fleet DNA drive cycle dataset is used to demonstrate these findings.
Download

Paper Nr:	10
Title:	A Two-stage Learning Approach for Traffic Sign Detection and Recognition
Authors:	Ying-Chi Chiu, Huei-Yung Lin and Wen-Lung Tai
Abstract:	With the progress of advanced driver assistance systems (ADAS), the development of assisted driving technologies is becoming more and more important for vehicle subsystems. The traffic signs are designed to remind the drivers of possible situations and road conditions to avoid traffic accidents. This paper presents a two-stage network to detect and recognize the traffic sign images captured by the vehicle on-board camera. In the detection network, we adopt Faster R-CNN to detect the location of the traffic signs. For the classification network, we use SVM, VGG, and ResNet for validation and testing. We compare the results and integrate the detection and classification systems. The datasets used in this work include TT100K and our own collected Taiwan road scene images. Our technique is tested using the videos acquired from the highway, suburb and urban scenarios. The results using Faster R-CNN for detection combined with VGG17 for classification have demonstrated superior performance compared to YOLOv3 and Mask R-CNN.
Download

Paper Nr:	12
Title:	Non-linear Motorcycle Dynamic Model for Stability and Handling Analysis with Roll Motion and Longitudinal Speed Regulation
Authors:	Vincenzo M. Arricale, Renato Brancati, Francesco Carputo, Antonio Maiorano and Guido Napolitano Dell’Annunziata
Abstract:	The use of computer simulations in motorcycle engineering makes it possible both to reduce designing time and costs and to avoid the risks and dangers associated with experiments and tests. The multi-body model for computer simulations can be built either by developing a mathematical model of the vehicle or by using commercial software for vehicle system dynamics. Even though the first method is more difficult and time-consuming than the second, maximum flexibility in the description of the features of the model can be obtained only by using an analytical model. Moreover, mathematical modelling has a high computation efficiency, whereas multi-body software requires a lot of time to carry out simulations. For the reasons above, the aim of this work was to develop a mathematical model of a motorcycle.
Download

Paper Nr:	14
Title:	A Systematic Approach of Reduced Scenario-based Safety Analysis for Highly Automated Driving Function
Authors:	Marzana Khatun, Michael Glaß and Rolf Jung
Abstract:	This paper investigates the scenario catalog generation and scenario reduction approaches for a complete Highly Automated Driving Function (HADF). Such approaches focus on the clustering and/or grouping of scenarios by applying a simple stochastic process at an early stage of development. Dealing with an enormous number of scenarios considering Functional Safety (FuSa), Safety Of The Intended Functionality (SOTIF) including cybersecurity desires intelligent approaches for HADF’s scenario reduction. The reduction of scenarios in HADF is a challenge for automotive researchers since it relates to a large number of parameters (like environmental aspects). The main contributions of the scenario generation and reduction approach proposed in this work are the following: (1) contribution to a complete scenario catalog for a dedicated HADF, (2) logical scenario optimization with parameter distribution, and (3) optimize discretization step for finding semi-concrete scenarios that can be executed. Furthermore, the optimization method incorporating the Monte-Carlo (MC) experiment with the CarMaker simulation yields a systematic approach to modeling reduced scenarios without redundancy to support safety.
Download

Paper Nr:	15
Title:	Autonomous Braking and End to End Learning using Single Shot Detection Model and Convolutional Neural Network
Authors:	Marwan Elkholy, Kirollos Nagy, Mario Magdy and Hesham H. Ibrahim
Abstract:	Safety issues concerning autonomous vehicles are becoming increasingly striking. Therefore, taking security issues of autonomous driving into account such as detection and identification of the vehicle in the surrounding is necessary to apply warning messages and braking based on the state of the vehicle. This paper develops an end to end deep learning, using different recognition algorithms, to promote the safety of autonomous vehicles in terms of controlling the steering and speed of a self-driving car. Two convolutional neural network architectures are presented with different number of filters in their layers. The networks were trained to take images as input data and scan the raw pixels and convert them directly into steering angle command and speed value. Also, an object recognition algorithm is provided which detects and determines the objects and their distances from the controlled car to have a collision warning system by using a pre-trained single shot detector model. All predicted speed values and steering angles, alongside the object detection model, are then translated into throttle and braking values while evaluating the models using a simulator and real road videos.
Download

Paper Nr:	20
Title:	Health Monitoring of Automotive Suspension System using Machine Learning
Authors:	Ahmed Abdelfattah and Hesham Ibrahim
Abstract:	This paper investigates Knowledge-based condition monitoring of automotive suspension dampers by implementing a quarter car model (QCM). The sprung mass acceleration - frequency power spectral density curves, for different cases of performance degradation in suspension damping and different operational conditions, is provided in response to the random road disturbance of different road classes. Training and testing acceleration response data are generated by Mtalb/simulink and fed to different classification algorithms that are trained and tested to distinguish between the different damping degradation values, in order to assess their performance in terms of classification accuracy as well as their confusion matrix. In addition, the worthiness of applying Principal Component Analysis (PCA), as a dimensional reduction technique, to increase all candidate classification algorithms is explored. Finally, the results of Quadratic Support Vector Machine showed the best performance in terms of accuracy and confusion matrix, while using dimensional reduction turned to be inefficient.
Download

Paper Nr:	33
Title:	Online State Estimation for Microscopic Traffic Simulations using Multiple Data Sources
Authors:	Kevin Malena, Christopher Link, Sven Mertin, Sandra Gausemeier and Ansgar Trächtler
Abstract:	The online fitting of a microscopic traffic simulation model to reconstruct the current state of a real traffic area can be challenging depending on the provided data. This paper presents a novel method based on limited data from sensors positioned at specific locations and guarantees a general accordance of reality and simulation in terms of multimodal road traffic counts and vehicle speeds. In these considerations, the actual purpose of research is of particular importance. Here, the research aims at improving the traffic flow by controlling the Traffic Light Systems (TLS) of the examined area which is why the current traffic state and the route choices of individual road users are the matter of interest. An integer optimization problem is derived to fit the current simulation to the latest field measurements. The concept can be transferred to any road traffic network and results in an observation of the current multimodal traffic state matching at the given sensor position. First case studies show promosing results in terms of deviations between reality and simulation.
Download

Paper Nr:	35
Title:	Analytical Approaches for Fast Computing of the Thermal Load of Vehicle Cables of Arbitrary Length for the Application in Intelligent Fuses
Authors:	Anika Henke and Stephan Frei
Abstract:	In modern intelligent vehicles, a huge number of components leads to complex cable harnesses with high reliability demands. Static connections protected by simple melting fuses are more and more replaced by intelligent power distribution and switching units. Thermal considerations play an important role with respect to reliability as thermal overload situations can lead to accelerated aging, damaged cables and finally to interruptions in the power supply. The calculation of the axial transient temperature distribution in cable structures is a complex task that is often solved numerically. In this paper, two analytical approaches to model the temperature of a single cable in air are presented, that are based on the use of Green’s functions in the time domain respectively Laplace domain. As sums appear, the convergence behavior is evaluated. The approaches are validated using a numerical reference solution. The influence of the cable length on the accuracy of the solutions is examined and complexity considerations are performed. An application example for intelligent vehicles is presented and discussed.
Download

Paper Nr:	36
Title:	The Perception Modification Concept to Free the Path of An Automated Vehicle Remotely
Authors:	Johannes Feiler and Frank Diermeyer
Abstract:	Inner-city, automated vehicles will face situations in which they leave their operational design domain. That event may lead to an undesired vehicle standstill. Consequently, the vehicle’s independent continuation to its desired destination is not feasible. The undesired vehicle standstill can be caused by uncertainties in object detection, by environmental circumstances like weather, by infrastructural changes or by complex scenarios in general. Teleoperation is one approach to support the vehicle in such situations. However, it may not be clear which teleoperation concept is appropriate. In this paper, a teleoperation concept and its implementation to free the path of an automated vehicle is presented. The situation to be resolved is that a detection hinders the automated vehicle to proceed. However, the detection is either a false positive or it is an indeterminate object which can be ignored. The teleoperator corrects the object list and the occupancy grid map. Thereby, the teleoperator enables the automated vehicle to continue its path. The preliminary tests show that the teleoperation concept enables teleoperators to resolve the respective scenarios appropriately.
Download

Paper Nr:	37
Title:	PolarNet: Accelerated Deep Open Space Segmentation using Automotive Radar in Polar Domain
Authors:	Farzan E. Nowruzi, Dhanvin Kolhatkar, Prince Kapoor, Elnaz J. Heravi, Fahed Al Hassanat, Robert Laganiere, Julien Rebut and Waqas Malik
Abstract:	Camera and Lidar processing have been revolutionized with the rapid development of deep learning model architectures. Automotive radar is one of the crucial elements of automated driver assistance and autonomous driving systems. Radar still relies on traditional signal processing techniques, unlike camera and Lidar based methods. We believe this is the missing link to achieve the most robust perception system. Identifying drivable space and occupied space is the first step in any autonomous decision making task. Occupancy grid map representation of the environment is often used for this purpose. In this paper, we propose PolarNet, a deep neural model to process radar information in polar domain for open space segmentation. We explore various input-output representations. Our experiments show that PolarNet is a effective way to process radar data that achieves state-of-the-art performance and processing speeds while maintaining a compact size.
Download

Paper Nr:	41
Title:	A Piecewise Linearization Algorithm for Solving MINLP in Intersection Management
Authors:	Matthias Gerdts, Sergejs Rogovs and Giammarco Valenti
Abstract:	In this paper, we propose a linearization algorithm for solving a Mixed Integer NonLinear Problem (MINLP) for Intersection Management (IM) of Connected Autonomous Vehicles (CAVs). The objective of such problem is to minimize the time it takes to clear a given arbitrary intersection for all vehicles in the consideration. We treat the IM problem as a bi-level optimization problem. On the lower level we solve an Optimal Control Problem (OCP) for each individual vehicle, whereas on the higher level we deal with an optimization problem of finding the optimal sequence and starting times for every car, which essentially yields a MINLP. An intuitive linearization technique is presented to solve the emerging MINLP in a reasonable time. The actual controls, if necessary, are computed a posteriori by minimizing the L2-norm of control variables. The algorithm is tested in different intersection scenarios. Numerical results show that it is suitable for real-time applications.
Download

Paper Nr:	45
Title:	Vegetation Detection in UAV Imagery for Railway Monitoring
Authors:	Md Atiqur Rahman and Abdelhamid Mammeri
Abstract:	Vegetation management on and alongside the railway tracks is very crucial for safe railway operations. The railway industry, therefore, needs to regularly monitor the growth of vegetation on railway tracks and embankments and mostly relies on human inspectors for the inspection and monitoring. This manual process being prohibitively time-consuming and cost-ineffective, there is a growing need to automate the process of vegetation detection. Aerial imagery collected using Unmanned Aerial Vehicles (UAVs) is becoming increasingly popular for automated inspection and monitoring. On the other hand, due to their recent success, Deep Convolutional Neural Networks (DCNNs) have seen rapid deployment in a wide array of image understanding tasks. In this work, we therefore, investigate the effectiveness of DCNNs for automating the vegetation detection task using UAV imagery. We further propose simple yet effective modification to an existing DCNN architecture and demonstrate its efficacy for vegetation detection using publicly available dataset.
Download

Paper Nr:	46
Title:	What Does Visual Gaze Attend to during Driving?
Authors:	Mohsen Shirpour, Steven S. Beauchemin and Michael A. Bauer
Abstract:	This study aims to analyze driver cephalo-ocular behaviour features and road vanishing points with respect to vehicle speed in urban and suburban areas using data obtained from an instrumented vehicle’s eye tracker. This study utilizes two models for driver gaze estimation. The first model estimates the 3D point of the driver’s gaze in absolute coordinates obtained through the combined use of a forward stereo vision system and an eye-gaze tracker system. The second approach uses a stochastic model, known as Gaussian Process Regression (GPR), that estimates the most probable gaze direction given head pose. We evaluated models on real data gathered in an urban and suburban environment with the RoadLAB experimental vehicle.
Download

Paper Nr:	47
Title:	Capturing the Variety of Urban Logical Scenarios from Bird-view Trajectories
Authors:	Christian King, Thilo Braun, Constantin Braess, Jacob Langner and Eric Sax
Abstract:	Driving scenarios are an essential part of validation of future highly automated driving (HAD) systems. In order to provide a valid proof of safety, it is crucial to test the system in as many realistic driving scenarios as possible. For this reason, it is necessary to extract driving scenarios from recorded data. A particular challenge in urban traffic is that there is a high degree of interaction between road users that needs to be considered. In this paper we present a concept for a maneuver-based extraction of driving scenarios. The extracted scenarios are provided in a format that supports a swift understanding of the content. In addition to the mere driving scenarios, parameter ranges for each scenario are grouped and aggregated from the data. Hence, the scenarios extracted with the presented concept can be used for re-simulation during the validation. We provide some results from the scenario extraction for an intersection from the INTERACTION data set.
Download

Paper Nr:	53
Title:	A Full-Featured, Enhanced Cost Function to Mitigate Motion Sickness in Semi- and Fully-autonomous Vehicles
Authors:	Isa Moazen and Paolo Burgio
Abstract:	Current full- and semi-Autonomous car prototypes increasingly feature complex algorithms for lateral and longitudinal control of the vehicle. Unfortunately, in some cases, they might cause fussy and unwanted effects on the human body, such as motion sickness, ultimately harnessing passengers' comfort, and driving experience. Motion sickness is due to conflict between visual and vestibular inputs, and in the worst case might causes loss of control over one’s movements, and reduced ability to anticipate the direction of movement. In this paper, we focus on the five main physical characteristics that affect motion sickness, including them in the function cost, to provide quality passengers' experience to vehicle passengers. We implemented our approach in a state-of-the-art Model Predictive Controller, to be used in a real Autonomous Vehicle. Preliminary tests using the Unreal Engine simulator have already shown that our approach is viable and effective, and we implemented and evaluated using Motion Sickness Dose Value and Illness Rating and then tested it in an embedded platform. We implemented it on our embedded platform, NVIDIA Jetson AGX Xavier that is representative of the next-generation AV Domain Controller.
Download

Paper Nr:	54
Title:	A Survey of UAS Technologies to Enable Beyond Visual Line Of Sight (BVLOS) Operations
Authors:	Elena Politi, Ilias Panagiotopoulos, Iraklis Varlamis and George Dimitrakopoulos
Abstract:	Latest trends, societal needs and technological advances have led to an unparalleled expansion in the use of Unmanned Aerial Systems (UAS) for versatile civilian and military applications, ranging from simple everyday operations, to the supervision in construction sites, even logistics, among others. Unmanned Aerial Vehicles (UAVs), widely known as drones, are the main components of UAS, and are becoming increasingly popular in such operations, since they reduce costs, they facilitate activities and can increase the granularity of surveillance or delivery. Furthermore, they can pave new ways for the implementation of smart-sensing and navigation functionalities, support automation, safety of operations, prognostics and even forensic analyses. Being an emerging technology, several challenges still need to be tackled in order to make UAS suitable for real-world applications, which impose strict performance, dependability and privacy requirements. In the light of the above, this paper provides an in depth survey of current UAS technologies for Beyond the Visual Line of Sight (BVLOS) UAS operations and highlight the main technological challenges and requirements that arise. We also focus on the emerging and future BVLOS UAS features and the technological advances that render their expansion in various industrial sectors promising.
Download

Paper Nr:	57
Title:	Let It Crash! Energy Equivalent Speed Determination
Authors:	Pavlína Moravcová, Kateřina Bucsuházy, Martin Bilík, Michal Belák and Albert Bradáč
Abstract:	Crash analysis including calculation of the impact speed and related determination of deformation energy is one of the main assumptions for the clarification of mostly negligent crimes. In this article were introduced results of two crash tests representing the comparison of the stiffness and technological obsolescence and their influence on the resulted vehicle deformation. Different extent of vehicle deformation was used to demonstrate the limits of selected methods for Energy Equivalent Speed determination as a value which expresses the kinetic energy dissipated by the vehicle during the contact phase.
Download

Paper Nr:	64
Title:	Colorimetric Space Study: Application for Line Detection on Airport Areas
Authors:	Claire Meymandi-Nejad, Esteban Perrotin, Ariane Herbulot and Michel Devy
Abstract:	We propose an adaptive color reference refinement process for color detection in an aeronautical application: the detection of taxiway markings based on images acquired from an aircraft. Road markings detection is a key functionality for autonomous driving, and is actively studied in the literature. However, few studies have been conducted on aeronautics. Road markings are often detected by using color priors, sensitive to perturbations. Color-based algorithms are still favored in this context as the markings color provides important information. Our proposed method aims at reducing the impact of weather conditions, shadowing and illumination variations on color-based markings detection algorithms. Our approach adapts a given color reference in order to define a new flexible yet robust color reference while maximizing its difference to other colors in the image. It is achieved through a statistical analysis of color similarity over a set of images, computed on several color spaces and distance functions, in order to select the most relevant ones. We validate our approach by analyzing the quantitative improvement induced by this method using two color-based markings detection algorithms, based on the Hough Transform and the Particle Filter.
Download

Paper Nr:	66
Title:	Evaluating Message Size of the Collective Perception Message in Real Live Settings
Authors:	Michael Klöppel-Gersdorf and Thomas Otto
Abstract:	The introduction of the Collective Perception Message (CPM) by ETSI offers new possibilities for connected and automated driving by enabling the exchange of object information between several participants. While this is surely beneficial, it also leads to higher load on the communication channel, which poses a problem, especially when considering IEEE 802.11p V2X communication. To overcome this problem, several mitigation strategies were formulated by ETSI. In the literature, several simulation studies regarding the effect on the communication can be found. Goal of this paper is to enrich the discussion with measurements from a real vehicle, showing how many objects might be available for CPM inclusion in the near to mid future.
Download

Paper Nr:	73
Title:	Generation of Road Reference Heading using GPS Trajectories for Accurate Lane Departure Detection
Authors:	Shahnewaz Chowdhury, Md. T. Hossain and M. I. Hayee
Abstract:	Lane departure warning system (LDWS) has significant potential to reduce crashes on roads. Most existing commercial LDWSs use image processing techniques with or without Global Positioning System (GPS) technology and/or high-resolution digital maps to detect unintentional lane departures. However, the performance of such systems is compromised in unfavourable weather or road conditions e.g., fog, snow, or irregular road markings. Previously, the authors proposed and developed an LDWS using a standard GPS receiver without any high-resolution digital maps. The previously developed LDWS relies on a road reference heading (RRH) of a given road extracted from an open-source low-resolution mapping database to detect an unintentional lane departure. This method can detect true lane departures accurately but occasionally gives false alarms i.e., it issues lane departure warnings even if a vehicle is within its lane. The false alarms occur due to the inaccuracy of RRH originated from inherent lateral error in open-source low-resolution maps. To overcome this problem, now authors propose a novel algorithm to generate an accurate RRH for a given road using a vehicle’s past trajectories on that road. The newly proposed algorithm to generate an accurate RRH for any given road has been integrated with the previously developed LDWS and extensively evaluated in the field to detect unintentional lane departures. The field test results show that the newly developed RRH generation algorithm significantly improves the performance of the previously developed LDWS by accurately detecting all true lane departures while practically reducing the frequency of false alarms to zero.
Download

Paper Nr:	75
Title:	A Flexible Scheduling Architecture of Resource Distribution Proposal for Autonomous Driving Platforms
Authors:	Hadi Askaripoor, Sina Shafaei and Alois Knoll
Abstract:	Autonomous driving has attracted a significant amount of attentions over the last ten years. Providing a flexible platform to schedule the executions of the tasks under hard real-time constraints is also a crucial matter which needs to be taken into account by the integration of intelligent applications. In this work, we propose a resource planner, consisting of a monitoring mechanism, context manager, and decision unit which facilitates the timing requirements in the presence of AI-based applications for the autonomous vehicles.
Download

Paper Nr:	84
Title:	Resolving Confusion of Unknowns in Autonomous Vehicles: Types and Perspectives
Authors:	Kaushik Madala and Hyunsook Do
Abstract:	Autonomous vehicles are susceptible to unknowns. In particular, vehicles with SAE level 5 of driving automation, which need to operate in complex operational design domain (ODD) conditions, have a very high chance to face unknowns. While the industrial standards ISO 21448 and UL 4600 hint at analyzing unknowns from the analysts and engineers’ perspective, the unknowns from different perspectives such as a autonomous vehicle or a machine learning model within an autonomous vehicle can differ from those perceived by engineers and analysts. In this paper, we discuss the different types of unknowns considering three different perspectives: analysts and engineers, autonomous vehicles, and machine learning (ML) models. We also clarify the often confused concepts of unknown knowns and unknowns unknowns for each perspective. Using a running example, we show how considering unknowns from different perspectives will aid in designing a safe autonomous vehicle.
Download

Paper Nr:	86
Title:	The Need for Location-based Machine Learning Models for Level 5 Automated Vehicles
Authors:	Kaushik Madala and Hyunsook Do
Abstract:	Assuring safety of machine learning (ML) models in autonomous vehicles is a challenging task. This is because of the complex operational design domain (ODD) settings under which we need to validate ML models. In particular, autonomous vehicles with level 5 of driving automation need to operate under any ODD conditions, and should ensure safety of both road users and passengers. However, deploying common ML models across a fleet of vehicles to operate in multiple regions can complicate the safety assurance process.Even when an ML model is found to be causing a crash due to an ODD condition occurring in only one of the regions, we still should update it across the fleets of all regions. If we can limit its update within that particular region, we can reduce the complexity of safety assurance. In this paper, we propose the location-based machine learning models for level 5 automated vehicles to address this problem and how they will be helpful compared to deploying instances of global ML models which are same across a company’s fleets of vehicles.
Download

Paper Nr:	92
Title:	Detection, Estimation & Tracking Road Objects for Assisting Driving
Authors:	Afnan Alshkeili, Wenliang Qiu and Bidisha Ghosh
Abstract:	The new era of mobility is moving towards automation. Detecting, estimating, and tracking objects from moving vehicles using dash-cam images in real-time can provide substantial advantages in supporting drivers’ decision making in advance. In this paper, an advanced deep learning-based object detection, distance estimation, and tracking framework has been proposed for this purpose. RetinaNet algorithm with ResNeXt backbone network has been used to detect five traffic object classes, including cars, cyclists, pedestrians, buses, and motorcycles, with improved accuracy. Additionally, distance estimation algorithm was introduced to increase both reliability and precession of detection. Moreover, an improved Simple Online and Real-time Tracking (SORT) algorithm were sequentially used to estimate traffic parameters such as volume and approach speed of each of these traffic object classes. The algorithm was trained and tested on stock imagery (COCO2017 and MOT16, respectively) of real-world videos taken from urban arterials with multimodal, signalized traffic operations.
Download

Paper Nr:	95
Title:	Safety-configuration of Autonomous Bus in Pedestrian Zone
Authors:	Qazi H. Jan and Karsten Berns
Abstract:	For self-driving vehicles to be equally trusted by the community like conventional vehicles and become a pivotal part of transportation, it is crucial to guarantee the safety of such vehicles. Safety must ensure that the vehicle will not collide with other obstacles and always stop in case of system failure. The vehicle used for the safety-configuration explained in this paper is a mini-bus that can carry around 10 passengers. It is intended to drive in a pedestrian-zone, an environment that involves many pedestrians and cyclists apart from occasional vehicles in a close-fitting space. Besides the manufacturer’s basic system provided to enable safety, safety certified system were added to trigger the safety at specific conditions. This includes emergency buttons, wireless safety system and configurable laser-scanners. This will allow the vehicle to stop based on physically activating the safety or automatically by laser-scanners. After various tests, the vehicle was able to brake immediately. This safety system is guaranteed not be influenced or disabled by any external system. This safety-configuration is to facilitate the entire system for safety-certification in the future.
Download

Paper Nr:	98
Title:	Study of Parameter Influence of the Basic Cylinder of Rotary Screw Propulsion Units on Noise Level during Locomotion on Ice
Authors:	Umar Vahidov, Dmitriy Mokerov, Roman Dorofeev, Vladimir Belyakov, Vladimir Makarov and Yuri Molev
Abstract:	The paper presents deformation calculations of the basic cylinder of the rotary screw propulsion unit under external load. The influence of such basic parameters of the rotary screw propulsion unit as basic cylinder diameter, its length and wall thickness on its deformation area, and, as a consequence, the generated noise level has been demonstrated. The basis for the method development were studies by scientists who were studying the sound wave generation by the deformation of different structural elements, parts and assemblies. The contribution of each of the parameters of the basic cylinder to the general level of the generated sound pressure was analyzed. It was determined that the magnitude of the noise level is mostly dependent on the length of the basic cylinder, then, to a less extent, on the cylinder wall thickness, and, to the least extent, on its diameter. A correlation was revealed between the oscillations power of the propulsion unit and their correlation with their generated noise level. The results and conclusions obtained during the described study allow for a more solid-based approach to the parameter selection of the rotary-screw propulsion unit for improved acoustic comfort, also for improved driver's work conditions, including the relief of his/her nervous system, sharpening of his/her attention during the operation, accident reduction. Beside the design of structures with the least possible noise, the proposed method allows also for selection of a rational propulsion unit design achieving a compromise between vehicle's noise specifications and its off-road ability.
Download

Paper Nr:	99
Title:	Towards a Rule-based Approach for Estimating the Situation Difficulty in Driving Scenarios
Authors:	Maximilian Schukraft, Susanne Rothermel, Juergen Luettin and Lavdim Halilaj
Abstract:	The task of safe driving poses a huge challenge for drivers in day to day driving situations. Many times, this task can be very difficult, e.g., due to dense traffic, bad weather conditions, or a risky driving maneuver, and thus demand high concentration of the driver. The difficulty level escalates by the ever-increasing infotainment offers inside vehicles or distractions caused by occupants thus making substantial contribution to the driver distraction. This often results in dangerous driving situations which could be avoided by Advanced Driver Assistance Systems or highly automated driving systems taking the situation difficulty into account. E.g., an incoming phone call is postponed during a difficult situation. However, current systems do not consider all factors that influence the difficulty of a given situation. In this paper, we present an approach for estimating the difficulty of a driving situation by combining a number of different factors, such as environmental, inside-vehicle, driver state and personal characteristics, respectively. Our approach follows a rule-based paradigm to make the difficulty estimation reproducible and adjustable to current traffic rules. It is based on a generic and modularized architecture to allow integration and abstraction from heterogeneous data sources. Further, a feedback is provided to the driver or system to explain the contribution of the various factors to the difficulty status. Finally, we demonstrate the capability of the proposed approach with concrete examples, where we estimate the difficulty in various driving scenarios and for different drivers.
Download

Paper Nr:	7
Title:	Road Traffic Anomaly Detection based on Deep Learning Technology
Authors:	Jamal Raiyn
Abstract:	Autonomous vehicles (AVs) collect big data based on various smart devices and sensors, with the goal of enabling a vehicle to operate under its own power. Fully automation vehicle is expected to have full control over all functions. Big data in intelligent transportation systems refers to large amounts of travel information. To manage large amounts of data in different formats collected via various types of communication channels, such as vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication, a deep learning concept is proposed, one that is inspired by the human central nervous system. Deep learning is a type of machine learning in which a model learns to perform classification tasks directly from input datasets. In this case, it also, refers to a three-layer structure: an input layer, a hidden layer, and an output layer. In this paper, various machine learning schemes have been proposed to detect anomalous conditions in urban road traffic. Furthermore, an evaluation of performance analysis based on simulation result of these schemes is performed. A deep learning concept is introduced to manage vehicle speed data, with the goal of detecting anomalous conditions in urban road traffic.

Paper Nr:	26
Title:	Collective Perception: Impact on Fuel Consumption for Heavy Trucks
Authors:	Juergen Hauenstein, Jakob Gromer, Jan C. Mertens, Frank Diermeyer and Sven Kraus
Abstract:	With on-board sensor technology, the environment can only be perceived to a limited extent. This can lead to energy-inefficient driving maneuvers due to the late perception of objects. The fuel consumption of heavy trucks is a major cost factor for transport companies, which is why energy-efficient systems are being sought. With collective perception, perceived objects are exchanged via Vehicle-to-Everything (V2X) and merged to a common environment model. Therefore, it is possible to achieve a greater awareness, which allows for improved planning for automated vehicles. In this publication, a system with collective perception and energy-efficient maneuver planning is presented. The functioning of the collective perception is presented using real vehicle data. A vehicle simulation shows the positive effect of collective perception in combination with an energy-efficient maneuver planner for determining the fuel consumption of heavy trucks.
Download

Paper Nr:	29
Title:	The Forerunner UAV Concept for the Increased Safety of First Responders
Authors:	Mihály Nagy, Péter Bauer, Antal Hiba, Attila Gáti, István Drotár, Balázs Lattes and Ádám Kisari
Abstract:	This paper proposes a novel Forerunner UAV concept to increase the safety of first responders by monitoring the road in front of their emergency ground vehicle (EGV) and notifying the driver about any violation of his/her right of way or approaching danger. The developments are conducted in an R&D project in Hungary. The proposed UAV for the planned urban demonstration is a hexacopter with triple redundant architecture applying a gimbaled camera to monitor the surroundings. In the cooperative control of the EGV and UAV the UAV must fly in front of the EGV which is possible through wireless communication of route data and velocity. Besides the real system a computer simulation representation is also applied including CARLA and Matlab to make exhaustive tests of the system capabilities. Increased attention is devoted to the possible wireless communication solutions as these are safety critical parts of the system. The article ends with the lists of planned simulation and real test scenarios to evaluate the system.
Download

Paper Nr:	44
Title:	Systems-theoretic Safety Assessment of Teleoperated Road Vehicles
Authors:	Simon Hoffmann and Frank Diermeyer
Abstract:	Teleoperation is becoming an essential feature in automated vehicle concepts, as it will help the industry overcome challenges facing automated vehicles today. Teleoperation follows the idea to get humans back into the loop for certain rare situations the automated vehicle cannot resolve. Teleoperation therefore has the potential to expand the operational design domain and increase the availability of automated vehicles. This is especially relevant for concepts with no backup driver inside the vehicle. While teleoperation resolves certain issues an automated vehicle will face, it introduces new challenges in terms of safety requirements. While safety and regulatory approval is a major research topic in the area of automated vehicles, it is rarely discussed in the context of teleoperated road vehicles. The focus of this paper is to systematically analyze the potential hazards of teleoperation systems. An appropriate hazard analysis method (STPA) is chosen from literature and applied to the system at hand. The hazard analysis is an essential part in developing a safety concept (e.g., according to ISO26262) and thus far has not been discussed for teleoperated road vehicles.
Download

Paper Nr:	50
Title:	Traffic Congestion “Gap” Analysis in India
Authors:	Tsutomu Tsuboi and Tomoaki Mizutani
Abstract:	This study is more than one-month traffic flow observation in India and introduces new traffic congestion “Gap” from the analysis of real traffic flow analysis in India. Traffic congestion becomes serious problem especially in developing countries such as India. In general, it is quite challenging to collect traffic data and understand traffic congestion problem from its data analysis. In this study, it is the first time to show long term traffic monitoring at one of major junction in Ahmedabad city of Gujarat state India. IAs for traffic congestion analysis, the following challenges are executed with a collaboration from local city government. Step 1 is to select location for 8 months observation by traffic monitoring camera in the city. Step 2 is to analyse traffic flow at the junction from each direction traffic flow. Step 3 is to evaluate traffic congestion with traffic flow parameter from traffic flow theory. Step 4 is to analyse geographical mapping by GIS tool. Based on these steps, it reached to unique traffic congestion mechanism in the junction, which it is named congestion “Gap” and large traffic volume is not always a case of traffic congestion. From this result, there is a possibility to improve traffic management when more detail observation at certain time of traffic congestion happing and environmental condition such as traffic signal control, road infrastructure structure and so on.
Download

Area 4 - Data Analytics

Full Papers

Paper Nr:	24
Title:	Time Series Segmentation for Driving Scenario Detection with Fully Convolutional Networks
Authors:	Philip Elspas, Yannick Klose, Simon Isele, Johannes Bach and Eric Sax
Abstract:	Leveraging measurement data for Advanced Driver Assistant Systems and Automated Driving Systems requires reliable meta information about covered driving scenarios. With domain expertise, rule-based detectors can be a scalable way to detect scenarios in large amounts of recorded data. However, rules might struggle with noisy data, large number of variations or corner cases and might miss valuable scenarios of interest. Finding missing scenarios manually is challenging and hardly scalable. Therefore we suggest to complement rule-based scenario detection with a data-driven approach. In this work rule-based detections are used as labels to train Fully Convolutional Networks (FCN) in a weakly supervised setup. Experiments show, that FCNs generalize well and identify additional scenarios of interest. The main contribution of this paper is twofold: First, the scenario detection is formulated as a time series segmentation problem and the capability to learn a meaningful scenario detection is demonstrated. Secondly, we show how the disagreement between the rule-based method and the learned detection method can be analyzed to find wrong or missing detections. We conclude, that the FCNs provide a scalable way to assess the quality of a rule based scenario detection without the need of large amounts of ground truth infromation.
Download

Paper Nr:	85
Title:	Feature-based Analysis of the Energy Consumption of Battery Electric Vehicles
Authors:	Patrick Petersen, Aya Khdar and Eric Sax
Abstract:	Battery electric vehicles have become increasingly important for the reduction of greenhouse gas emission. Even though the number of battery electric vehicles is increasing, the general acceptance and widespread introduction to consumers is still related to smaller range, which is in part due to the range anxiety leading to inefficient usage of the complete battery. Thus, an accurate range estimation is a key parameter for increasing the trust in the promised range, but accurate estimation is a nontrivial task. Advanced algorithms estimate the energy consumption based on the travel route and other non-deterministic factors such as driving style, traffic and weather conditions. The possible feature space is huge, therefore, the identification of a few highly energy consumption relevant features is necessary due to time and memory limitations in the vehicle including the improvement of the estimation itself. In this paper we present a data-driven methodology for systematically analyzing and engineering relevant features which influence the energy consumption concurrently, covering not only the driver style but also features based on road topology, traffic and weather conditions. Utilizing a real-world data set different trip segmentation methods and feature selection algorithms are compared to each other in regards to their accuracy and time-efficiency.
Download

Short Papers

Paper Nr:	38
Title:	Car Drivers Do Not Choose Their Speed in Urban Environments: Speed Models in Tangent Streets
Authors:	Yasmany García-Ramírez, Luis Paladines, Christian Verdesoto and Patricio Torres
Abstract:	The performance-based design approach is one way to deal with speeding in the streets. Under this approach, the geometric elements of roadways can influence on the desired operating speeds. Thus, several studies have investigated the relationship between geometric elements and light vehicle speeds; however, no conclusive results have been reached at this stage. In this context, this article aims to investigate the influence of several characteristics from urban street tangents, car driver, and vehicle on their speed in free-flow conditions. Three tangents scenarios were set out: before stop-controlled intersections, before signal-controlled intersections, and before roundabout intersections. Speeds of light vehicles were measured at 34 streets. Speeds were collected with in-vehicle GPS equipment. Thirty-five car drivers participated in the study with their vehicles. Street geometric characteristics, street environment variables, driver and vehicle characteristics were also collected. As a result, 15 regression models were calibrated and validated. Street length and objects density were the most influential variables in those models, and not the driver and vehicle characteristics as would suppose. This comprehensive research extends the knowledge of the most influential variables on speed in several urban scenarios, offering useful information for urban planners and street designers.
Download

Paper Nr:	88
Title:	Explainable Federated Learning for Taxi Travel Time Prediction
Authors:	Jelena Fiosina
Abstract:	Transportation data are geographically scattered across different places, detectors, companies, or organisations and cannot be easily integrated under data privacy and related regulations. The federated learning approach helps process these data in a distributed manner, considering privacy concerns. The federated learning architecture is based mainly on deep learning, which is often more accurate than other machine learning models. However, deep-learning-based models are intransparent unexplainable black-box models, which should be explained for both users and developers. Despite the fact that extensive studies have been carried out on investigation of various model explanation methods, not enough solutions for explaining federated models exist. We propose an explainable horizontal federated learning approach, which enables processing of the distributed data while adhering to their privacy, and investigate how state-of-the-art model explanation methods can explain it. We demonstrate this approach for predicting travel time on real-world floating car data from Brunswick, Germany. The proposed approach is general and can be applied for processing data in a federated manner for other prediction and classification tasks.
Download

Paper Nr:	62
Title:	Ambulance Vehicle Routing under Pandemic with Fuzzy Cooperative Game via Smart Contracts
Authors:	Alexander Smirnov and Nikolay Teslya
Abstract:	The pandemic caused by COVID-19 virus has posed a challenge for healthcare systems in many countries. One of the important tasks facing after a sick person detection is the timely patient’s transportation to a hospital. When making a decision on transportation to the hospital, it is necessary to account for many parameters, including beds space availability, availability of hospital staff and medicines required by the treatment protocol, diagnostic equipment on ambulance team, the distance to the hospital, ambulance vehicle locations, as well as hospital and ambulance staff psychophysical state, and patient’s reaction to hospitalization. Some of them can be gathered through smart city sources, like city databases or operational systems, but most of them require access to medical services. It is proposed to consider hospitals as participants of a cooperative game, whose overall goal is to ensure the maximum of cured patients. To describe the psychophysical state of the personnel, as well as to ensure greater variability of the resulting solution, the game parameters are proposed to be set using fuzzy sets and fuzzy logic. To implement the game rules, it is proposed to use smart contracts in blockchain technology. The blockchain could also be used to provide access to data from medical services, store and distribute the current state of hospitals, and save processing results for later analysis and model refinement.
Download

Area 5 - Smart Mobility and Sustainable Transport Services

Full Papers

Paper Nr:	17
Title:	Total Cost of Ownership for Automated and Electric Drive Vehicles
Authors:	Lambros Mitropoulos, Konstantinos Kouretas, Konstantinos Kepaptsoglou and Eleni Vlahogianni
Abstract:	Advances in technology and alternative fuels change the on-road vehicle fleet mix, which traditionally depends on internal combustion vehicles. These changes affect also the total cost of ownership (TCO) per vehicle technology and their market penetration rates. The goal of this paper is to identify indicators for a TCO based analysis for three vehicle technologies: A Hybrid Electric Vehicle (HEV), an Electric Vehicle (EV) and an Automated Electric Vehicle (AEV). The study is conducted by using data for the French market, for existing vehicle models; thus, the level three or “conditional driving automation” is used for the AEV. The assessment shows that while the EV is the most economical vehicle when considering the TCO, the HEV is more economical during the first two years. The high purchase cost of the AEV does not compensate during the vehicle lifetime compared to the other two technologies, although it profits from lower maintenance and time costs. The HEV approximates the AEV TCO at the end of its lifetime, however the higher expected resale value of the HEV make it attractive for consumers that desire lower purchase cost and higher resale value.
Download

Short Papers

Paper Nr:	16
Title:	Functional Safety and Electric Vehicle Charging: Requirements Analysis and Design for a Safe Charging Infrastructure System
Authors:	Tommi Kivelä, Mohamed Abdelawwad, Marvin Sperling, Malte Drabesch, Michael Schwarz, Josef Börcsök and Kai Furmans
Abstract:	As society moves from fossil fuels towards electric mobility, there’s an increasing need for charging infrastructure for electric vehicles. Aside from a network of public charging stations, charging equipment will also be increasingly installed in homes and used on a daily basis to charge electric vehicles (EVs) overnight. With this increasing role of charging infrastructure in day-to-day life, safety and security should be guaranteed for these systems. In this work we present the requirements analysis and a charging infrastructure system design for both private and public charging stations, with the goal of fulfilling the requirements of current functional safety and EV supply equipment (EVSE) standardization. Risk assessment for the charging process and the derived functional safety requirements are presented. The overall system design is discussed, with the main focus on the safety-related parts. The presented work can be used as a basis for the development of functionally safe next generation EVSE.
Download

Paper Nr:	87
Title:	Wireless Power Transfer with Data Transfer Capability for Electric and Hybrid Vehicles: State of the Art and Future Trends
Authors:	Sami Barmada, Nunzia Fontana and Mauro Tucci
Abstract:	This papers shows how Powerline Communication and Wireless Power Transfer technologies can be used together to allow both power and data transfer when hybrid and electric vehicles are connected to the grid. These two technologies have lately become popular when dealing with the Smart Grid environment (the former) and charging of electric powered devices (the latter). The authors have dedicated their research on the integration between them, keeping in mind their use in the automotive environment; this papers serves as a review and a starting point for future work in the area, offering a synthetic description of the operating principles and some results. In addition, shielding techniques for Wireless Power Transfer systems are shown and compared with each other, in order to show different aspects of this fundamental topic.
Download

Paper Nr:	30
Title:	User Experience and Analysis of an Autonomous Shuttle Service
Authors:	Lova Andersson, Allegra Ayala, Shuan Chan, Kyle Hickerson, Liam Kettle, Lindsey A. Malcein and Yi-Ching Lee
Abstract:	As the use of autonomous vehicles for public transportation becomes more prevalent, it is important to examine characteristics of potential users and their perception of the service. This study aimed to capture user opinions and feedback from both riders and non-riders concerning an autonomous shuttle service. Potential differences in user groups were examined as well, comparing employees of the Department of Defense to civilian users. Participants generally held positive opinions about the shuttle, although riders were more likely to rate the service favourably. Civilian users were also more likely to rate the shuttle favourably and more often claimed that they would recommend it to others. The youngest participants tended to report higher levels of agreement and acceptance on perceived safety and intelligence as well as the shuttle’s avoidance of obstacles and obedience of traffic rules. Research in this area has implications for all facets of the transportation industry as well as future users of autonomous public transportation.
Download

Paper Nr:	71
Title:	A Review on Charging Systems for Electric Vehicles in Smart Cities
Authors:	Mohamed A. El Ghany
Abstract:	An overview of types of electric vehicles and the variant batteries which the electric vehicles currently use is demonstrated. Different charging systems are presented. Plug-in charging system which is either on-board or off board is investigated. Moreover, fundamentals of wireless charging are analyzed for smart cities. Current market perspectives are provided. A hybrid charging system is also discussed. Different approaches for vehicle to vehicle charging either using plug or wireless charging are provided.
Download

Paper Nr:	93
Title:	A Novel Approach of Environment Impact Assessment and Emission Measurement on the Inter-city Transportation in the Greater Bay Area (GBA) of China using a Modified Gravity Model
Authors:	Eugene C. Wong, Danny K. Ho, Stuart So, Eve H. Chan and Chi-Wing Tsang
Abstract:	The Guangdong-Hong Kong-Macau Greater Bay Area (GD-HK-MO) also referred as Greater Bay Area (GBA), is a megalopolis, consisting of nine cities and two special administrative regions, i.e., Hong Kong and Macao in South China. GBA has a total population of approximately 71.2 million people representing about 5% of China’s total population with a combined regional GDP at USD 1642 billion in 2018, i.e. about 12% of GDP for the whole mainland China. Hong Kong acting as a window of China, plays a critical role in contributing to the growth of the GDP. Given the enormous scale of this regional economy and increasing collaboration among these GBA cities, it is utmost important to design a novel environmental impact assessment and emission measurements of the cross-border transportation among Hong Kong and various GBA cities with the aim of proposing countermeasures on carbon emissions of vehicles in the transport and logistics sector of the GBA. In the study, two modified gravity models are designed by considering social, economic, and other variables affecting the carbon emission of vehicles travelling within and across cities in the GBA. Further study will be pursued using decomposition analysis based on the modified gravity model to analyse the crucial contributors and determinants of carbon emission among the GBA cities.
Download