Archives Search

Abstract

The service of Virtual Infrastructure Simulation and Evaluation (VISE) offers evaluation of road sections with “human in the loop” procedure. It consists of virtual infrastructure simulation of planned (not yet built) road designs tested by end users using a driving simul...Lead moreThe service of Virtual Infrastructure Simulation and Evaluation (VISE) offers evaluation of road sections with “human in the loop” procedure. It consists of virtual infrastructure simulation of planned (not yet built) road designs tested by end users using a driving simulator in early design stages. This enables optimization of cost and performance of planned road sections, since the cost of design changes is much lower in early design phases. Furthermore, VISE can also be used for evaluation of existing road sections and proposals for their optimisation, by visually simulating and testing any potential changes and improvements. With this service, real drivers test and evaluate the newly designed road section in a high-fidelity motion-based driving simulator, which represents a safe, yet equally engaging driving environment. An important part of this solution is also the digitalisation and virtualization of new road sections using Building Information Modelling (BIM).

Abstract

Operational Design Domains or ODDs crucially determine the parts of the road and street network, and the times, when vehicles can be used in the automated mode. ODDs depend on the capabilities of the automated vehicles, and especially on their sensors and artificial intel...Lead moreOperational Design Domains or ODDs crucially determine the parts of the road and street network, and the times, when vehicles can be used in the automated mode. ODDs depend on the capabilities of the automated vehicles, and especially on their sensors and artificial intelligence. These capabilities will likely dramatically increase, and their prices decrease during the next decades, increasing ODD continuity and coverage, and thereby HAV acceptance. The paper proposes a generic list of ODD features, describes the current ODDs for five highly automated driving use cases as well as the current unit costs of the infrastructure elements facilitating the ODDs. The high cost elements include HD mapping including LIDAR point clouds, low-latency broadband connectivity, enhanced snow removal, access control furniture, provision of safe harbours along roads, VMS or C-ITS warnings, and real time situational picture.

Abstract

The level of automation in vehicles will significantly increase over the next decade. As automation will become more and more common, vehicles will however not be able to master all traffic related situations for a long time. In such situations, the driver must take over ...Lead moreThe level of automation in vehicles will significantly increase over the next decade. As automation will become more and more common, vehicles will however not be able to master all traffic related situations for a long time. In such situations, the driver must take over and steer the vehicle through the situation. One of the important questions is when the takeover should be performed. There are many decisive factors that must be considered. On the one hand, the current traffic situation including roads, traffic light and other road users, especially vulnerable road users, and on the other hand, the state of the driver must be considered. The goal is to combine neurocognitive measurement of the drivers’ state and the static and dynamic traffic related data to develop an interpretation of the current situation. This situation analysis should be the basis for the determination of the best takeover point.

Abstract

This paper investigates the level of investments in digitalisation in the transport industry and compares it with other industries. The two countries covered by this study are Australia and Finland, both fairly advanced in digitalisation. The data used for the analysis is...Lead moreThis paper investigates the level of investments in digitalisation in the transport industry and compares it with other industries. The two countries covered by this study are Australia and Finland, both fairly advanced in digitalisation. The data used for the analysis is extracted from the national accounts. The level of digitalisation is approximated by the investments in information and communications technology. The results show that transport industry is not at all in the leading in investing in and adoption of new technology, but rather the contrary. Industries, companies and organisations promoting and deploying ITS have a substantial challenge to change the current patterns of low investments. The paper discusses potential reasons for these patterns, and the policies and strategies that are called for.

Abstract

Development and deployment of Cooperative Intelligent Transport Systems (C-ITS) take place with the aim to improve safety, traffic efficiency, energy efficiency and comfort. The ambition of the European Commission and the EU Member States at this stage is to initiate and ...Lead moreDevelopment and deployment of Cooperative Intelligent Transport Systems (C-ITS) take place with the aim to improve safety, traffic efficiency, energy efficiency and comfort. The ambition of the European Commission and the EU Member States at this stage is to initiate and establish large-scale deployment of sustainable services on all road categories with the support of authorities at different levels, and to ensure interoperability, security, and seamless availability of high-quality services for end-users. This paper explores operational procedures for large-scale C-ITS deployment in Europe, and targets two groups of authorities, road operators and municipality members, as well as technical developers for the implementation of C-ITS services. The paper also provides guidelines with technical details of the C-ITS services, such as Day-1 and Day-1.5 services, as defined by the EU C-ITS Platform for properly implementing pan-European interoperable C-ITS services.

Abstract

ITS is traditionally about optimizing traffic flow, reducing the numbers of stops, and increasing speeds for cars. However, in support of the increasing interest of cities around the world to accelerate cycling as transport in cities ITS solutions to optimize flow, reduce...Lead moreITS is traditionally about optimizing traffic flow, reducing the numbers of stops, and increasing speeds for cars. However, in support of the increasing interest of cities around the world to accelerate cycling as transport in cities ITS solutions to optimize flow, reduce the number of stops, and increase safety for people on bicycles have been developed and implemented. The paper will present some of the different ITS solutions like green waves for cyclists, rain sensors, cyclist counters as well as way finding and safety solutions implemented in Denmark and make the case that such solutions could be easily implemented elsewhere to achieve common goals of better mobility, better use of public space, congestion reduction, better quality of life for citizens, better public health, a more vibrant urban life, and reduced noise, CO2-emissions, and pollution.

Abstract

Due to strict regulations by the European Commission on rest periods for truck drivers, they are forced to rest regularly. A significant problem arises when a truck parking area is completely full that trucks are parked in inconvenient and dangerous places, e.g. on- and o...Lead moreDue to strict regulations by the European Commission on rest periods for truck drivers, they are forced to rest regularly. A significant problem arises when a truck parking area is completely full that trucks are parked in inconvenient and dangerous places, e.g. on- and off-ramps. We propose a solution to supply truck drivers with predictions of future occupancy levels to enable them to optimally plan their rest periods. A novelty of the proposed solution is that it uses not only temporal and occupancy level information but also traffic and environmental data from a diverse sensor network along the highways and at the concerned parking areas. We show that the prediction accuracy is significantly higher when accounting for a more comprehensive dataset. The validation results show that it is possible to generate such predictions with relatively high accuracy for a time range of up to five hours ahead and that the predictions are most accurate in the late afternoon when truck drivers are most likely to use the prediction service. Our approach is also highly relevant for adjacent application fields such as Park+Ride areas connecting individual traffic with public transit.

Abstract

The automotive industry is committed to satisfy stricter regulations of energy demand and emission. Simulation-based approaches are gaining more importance to investigate such problems and initiate countermeasures at an early stage. Microscopic traffic simulation and driv...Lead moreThe automotive industry is committed to satisfy stricter regulations of energy demand and emission. Simulation-based approaches are gaining more importance to investigate such problems and initiate countermeasures at an early stage. Microscopic traffic simulation and drivetrain simulation are both discipline-specific simulation tools. This study introduces a generic validation approach to combine them together and provide an effective, realistic and cost-efficient co-simulation platform. In this regard, the district of Aachen is selected for the traffic simulation in which a drivetrain model is integrated. Results indicate that the proposed generic approach can effectively validate the complex traffic network and verify the drivetrain model.

Abstract

When considering daily activities a new approach is presented by introducing spatial and temporal flexibility of activities. The activities of the travellers are prioritized and ordered by solving the Traveling Salesman Problem. With the introduction of flexibility new lo...Lead moreWhen considering daily activities a new approach is presented by introducing spatial and temporal flexibility of activities. The activities of the travellers are prioritized and ordered by solving the Traveling Salesman Problem. With the introduction of flexibility new locations of the activities are suggested to the travellers, so that the optimization can be realized and shorter travel times can be achieved. Real activity chains were analyzed to assess the benefits of the optimization algorithm. The benefits were measured in terms of the average travel time saving per person per day. The evaluation was run on a test dataset and then it was generalized to a small and a big dataset. The average travel time saving for selected activity chains was 21,6%, where an optimization was feasible, while in general considering all activity chains it was 6,7%.

Abstract

Machine learning (ML) software is increasingly being used within safety-critical systems being developed for automated vehicles. The inclusion of ML software components improves the overall performance of the software but makes the elaboration of the system safety argumen...Lead moreMachine learning (ML) software is increasingly being used within safety-critical systems being developed for automated vehicles. The inclusion of ML software components improves the overall performance of the software but makes the elaboration of the system safety argument more challenging as the normal processes and methods are no longer applicable or sufficient. Some existing processes can be modified, but new argument structures and supporting methods need to be developed. This paper examines some of the challenges and proposes some new approaches to handle the safety analysis of ML software components.

Abstract

Public transport fleet operators are interested in lowering emissions due to tightened standards for city air quality. However, stakeholders in the electrification process with battery electric buses (BEB) are exposed to a complex network of dependencies from governmental...Lead morePublic transport fleet operators are interested in lowering emissions due to tightened standards for city air quality. However, stakeholders in the electrification process with battery electric buses (BEB) are exposed to a complex network of dependencies from governmental, economic, transport, and energy domains. Particularly, specialized tools for planning routes, vehicles, and charging infrastructure are missing at municipal-operational level of BEB deployment. Aim of the project as core of this paper is to develop methods and tools to support planning aspects of bus fleet electrification. Planners should be enabled to decide about the feasibility of deploying BEB on selected routes within the public transportation network by using input data of network design, schedule, vehicle characteristics, and charging infrastructure. Using the developed vehicle simulation model of BEBs, it is possible to scale battery capacities for full day duty-cycles on real bus routes under consideration of terrain grades, realistic traffic conditions, and opportunity charging.

Abstract

The existing and free of charge ASFINAG App is seen as a midterm bridge between today’s classic information services and the future mobility dominated by connected vehicles interacting with a cooperative road infrastructure. In 2017 the ASFINAG App has been extended with ...Lead moreThe existing and free of charge ASFINAG App is seen as a midterm bridge between today’s classic information services and the future mobility dominated by connected vehicles interacting with a cooperative road infrastructure. In 2017 the ASFINAG App has been extended with the launch of a feature called “Companion”: An information service with special focus on usage while driving on the motorway. To ensure safe usage dedicated design principles have been elaborated and introduced. The paper will present current state of play (features, content…) as well as an outlook to future developments.

Abstract

Autonomous Emergency Braking Systems have the potential to improve safety for pedestrians and cyclists. PROSPECT (Proactive Safety for Pedestrians and Cyclists) is a collaborative research project funded by the EC. Its objective was to improve effectiveness of active VRU ...Lead moreAutonomous Emergency Braking Systems have the potential to improve safety for pedestrians and cyclists. PROSPECT (Proactive Safety for Pedestrians and Cyclists) is a collaborative research project funded by the EC. Its objective was to improve effectiveness of active VRU safety systems compared to those currently on the market by: expanding the scope of scenarios addressed, improving the AEB/AES system performance and proposing extensive validation methodologies. All relevant VRU traffic scenarios were considered with a special focus on urban environments. The project comprises accidents analysis combined with results from naturalistic observation and HMI guidelines. This analysis represented key input for the system specifications, integration and demonstration with a cost-benefit analysis. For system development, two main aspects were considered: advanced sensor processing and actuator strategies (braking and/or steering). The concepts for sensors and control systems were shown in three vehicle demonstrators and one driving simulator. Each demonstrator addressed barriers of current AEB systems such as sensors field-of-view, path prediction, intent recognition and reaction times for the actuation. These were extensively tested in realistic scenarios with novel pedestrian and cyclist dummies including propulsion system. PROSPECT test methodologies and tools shall be considered for 2020 - 2024 Euro NCAP test programmes.

Abstract

Safe, environmentally conscious and flexible - these are the central requirements for the mobility of the future. In the border region between Germany, France and Luxembourg, mobility in the world of work and pleasure is a decisive factor. It must be simple, affordable an...Lead moreSafe, environmentally conscious and flexible - these are the central requirements for the mobility of the future. In the border region between Germany, France and Luxembourg, mobility in the world of work and pleasure is a decisive factor. It must be simple, affordable and available to all. The automation and intelligent networking of road traffic plays an important role in this. Due to the distributed settlement structure with many small towns and village and a few central hot spots, a fully available public transport is very complex and expensive and only a few bus and train lines exist. In this context, the tri-national research project TERMINAL (Automated electric minibuses in cross-border commuter traffic) aims to establish a cross-border automated minibus in regular traffic and to explore the user acceptance for commuter traffic. Additionally, mobility-on-demand services are tested, and both will be embedded within the existing public transport infrastructure.

Abstract

The first generation of C-ITS – often called “Day 1” services – is ready for deployment, inter-operable between different road operators and vehicle manufacturers thanks to extensive harmonization work involving all stakeholders. Work on the next generation of C-ITS servi...Lead moreThe first generation of C-ITS – often called “Day 1” services – is ready for deployment, inter-operable between different road operators and vehicle manufacturers thanks to extensive harmonization work involving all stakeholders. Work on the next generation of C-ITS services has already started: ASFINAG, the Austrian road operator, together with ITS industry partners has specified and freely published a set of eleven "Day 2" C-ITS services with the specific aim of supporting automated driving. The services deal with topics like route clearance, distance gaps, platoon support, vehicle type and/or lane specific speed recommendation but also broader issues like collective perception of objects or persons on the road or GNSS enhancement. All services are based on already standardized or pre-standardized C-ITS message formats. The paper introduces the services and their message formats and states the road ahead to transform the current specification into fully implemented, tested and harmonized services in the C-ITS domain.

Abstract

Abstract The forthcoming highly automated vehicles (HAVs, SAE Level 3) would introduce a new type of human-machine interaction that allows the drivers to be completely disengaged from driving and safety perform other non-driving related tasks. However, there are still sit...Lead moreAbstract The forthcoming highly automated vehicles (HAVs, SAE Level 3) would introduce a new type of human-machine interaction that allows the drivers to be completely disengaged from driving and safety perform other non-driving related tasks. However, there are still situations where the HAV system would still need human drivers to take over control back within a sufficient lead time. In HAVs, drivers may not only perform a single task but also engage in multiple tasks concurrently. This study aims to investigate the effects of engaging in multitasking on drivers’ takeover performance in HAVs. A driving simulator investigation with 8 participants (aged 20-49 years, 4 females and 4 males) was implemented. Results showed that multitasking leads to prolonged takeover time and slowed decision-making. The findings emphasize the importance of including end-users in the design process of HAVs.

Abstract

The below article describes the experience of the Republic of Tatarstan implemented by the Мinistry of informatization and communications of the Republic of Tatarstan in terms of implementation and development of intellectual transport systems elements intended for the co...Lead moreThe below article describes the experience of the Republic of Tatarstan implemented by the Мinistry of informatization and communications of the Republic of Tatarstan in terms of implementation and development of intellectual transport systems elements intended for the coordination of emergency services providing safety of passenger transportation applying GLONASS-based satellite technologies. In order to achieve the maximum efficiency of the interdepartmental interaction, the Мinistry of informatization and communications of the Republic of Tatarstan has developed and implemented the GLONASS+112 project – a comprehensive solution, the only one in the Russian Federation that enables to monitor the means and resources, to handle telephone calls from sufferers and route such calls between various departments of emergency services automatically, to enter and handle unified account cards for accidents for all the emergency services engaged in the accident response activities as a part of the unified interdepartmental information environment.

Abstract

Rapidly evolving IT solutions such as machine learning and database innovations encourage transport industry to look at modern technologies as an efficient cost reduction instrument. Still, business is burdened by complicated systems, delays and outdated document correspo...Lead moreRapidly evolving IT solutions such as machine learning and database innovations encourage transport industry to look at modern technologies as an efficient cost reduction instrument. Still, business is burdened by complicated systems, delays and outdated document correspondence. Blockchain is seen as a long-term solution that can manage uncertainties in document paper flow and in goods tracking. The aim of this research is to define key scenarios among on-going blockchain commercial and scientific projects within maritime port management. The research reviews how distributed databases can influence on port’s inner-organization, paper flow and cost reduction, actor’s cooperation and other parameters. A review of current projects and concepts is presented. Three key scenarios are defined that all reviewed projects could be compiled to: document workflow management, financial processes and device connectivity. Reviewed projects show a clear tendency to link blockchain and device connectivity, integrating technologies into current data management systems.

Abstract

This paper will present the first realistic simulation of on-demand mobility services with extensive Key Performance Indicator (KPI)-based analysis within well-defined service typologies. It will show how sophisticated matching, dispatching, and routing algorithms can be ...Lead moreThis paper will present the first realistic simulation of on-demand mobility services with extensive Key Performance Indicator (KPI)-based analysis within well-defined service typologies. It will show how sophisticated matching, dispatching, and routing algorithms can be tailored to improve fleet utilization, passenger service levels, or a weighted combination of both. Multiple scenarios will be tested and explored to demonstrate how planners can balance trade-offs between passenger comfort and convenience and operator requirements. Existing literature on mobility service design and simulation will be reviewed. Then the paper will describe the design and analysis of micro-transit, ride sharing, and shuttle on-demand mobility services in three service areas in the U.S. city of Chicago as examples of how planners can use simulation to optimize service performance before deployment to reduce risks of failure. The study will use Chicago taxi data to compare how shared on-demand services perform compared to actual unshared taxi service in three service areas. Passenger KPIs measured include excess ride time, and pick-up time deviation caused by sharing services and operator KPIs such as trip acceptance rates, vehicle occupancy, kilometers traveled, and effective kilometers traveled. The implications of the shared service on congestion and pollution will be touched upon.

Abstract

Austria’s major traffic infrastructure and traffic service providers established a nationwide intermodal traffic information platform called VAO that integrates approved traffic information of all modes of transport (incl. road, rail, cycling, walking) provided by the aut...Lead moreAustria’s major traffic infrastructure and traffic service providers established a nationwide intermodal traffic information platform called VAO that integrates approved traffic information of all modes of transport (incl. road, rail, cycling, walking) provided by the authorities. In case of public transport information this information is based on timetable and real time data (including delays etc.) for various operators. VAO has successfully implemented e.g. high-quality intermodal routing that is used by partners and b2b customers to create powerful end user applications. To enhance the real time road traffic information for the platform, as well as provide a countrywide authorized data basis for traffic information, traffic management and traffic analysis, Austria’s Ministry of Transport, Innovation and Technology (BMVIT) initiated a project, led by ASFINAG, to develop a harmonized real time traffic information network through its climate and energy fund (EVIS). The paper reports on objectives and status for these platforms.

Abstract

The basic assumption which led to this research was, that, despite it is often mentioned that MaaS is a user-centric approach, the human nature of bounded rationality and the society´s impact on individual mobility behaviour has not been considered adequately so far. Ther...Lead moreThe basic assumption which led to this research was, that, despite it is often mentioned that MaaS is a user-centric approach, the human nature of bounded rationality and the society´s impact on individual mobility behaviour has not been considered adequately so far. There are, amongst others, two main non directly observable psychological variables which massively influence mobility behaviour and the acceptance of MaaS: habit on the one hand and control beliefs on the other hand. Moreover this research suggests that it is also crucial, considering an individually perceived added value of multimodal mobility services as an important element towards their success. This paper wants to give some input and explanation on how the rational concept of MaaS could be elaborated and developed by the findings of social and psychological sciences to better meet the consumer needs and to strengthen its capacities to modernize travel behaviour in urban areas.

Abstract

In times where everyone speaks about autonomous driving, vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication plays an important role. Several pilot project had been deployed throughout Europe in order to prove the maturity of the technology, while t...Lead moreIn times where everyone speaks about autonomous driving, vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication plays an important role. Several pilot project had been deployed throughout Europe in order to prove the maturity of the technology, while the interoperability between various suppliers still needs to be proved. Slovenian national motorway operator DARS implemented its first V2I pilot system on a 24-kilometer-long motorway section including V2I road side units, central C-ITS back office and has distributed a few V2I on-board units to testing vehicles in order to prove the performance of the system. Several scenarios were tested including manually triggered one’s from the Traffic Control Center operator (traffic jam, accident, road works, obstacle on road) and automatically triggered scenarios receiving the information from external data sources about severe weather events and ghost driver detection. In the opposite direction the central system was receiving the data collected from vehicles.

Abstract

This document presents some methods proposed for the IEEE NGV (NEXT GENERATION V2X) task group, sometimes also referred to as IEEE 802.11px or IEEE 802.11bd, which is designated to be newest standard for V2X applications based on IEEE access layer. We describe why NGV pro...Lead moreThis document presents some methods proposed for the IEEE NGV (NEXT GENERATION V2X) task group, sometimes also referred to as IEEE 802.11px or IEEE 802.11bd, which is designated to be newest standard for V2X applications based on IEEE access layer. We describe why NGV proposals must ensure compatibility with the already deployed IEEE 802.11p networks, and how they can improve performance for new applications while guarantying fairness between all stations. Such new methods include adaptive retransmissions scheme based on channel state, and the appending of new-modulation symbols after the legacy PPDU with adaptive modulation and coding rate.

Abstract

This paper treats the dilemma between phase change prediction and phase change control by vehicle actuation. Traffic dependency is the greatest enemy of predictability. Intersections with cyclic control are easily predictable, even from the outside by observing its phases...Lead moreThis paper treats the dilemma between phase change prediction and phase change control by vehicle actuation. Traffic dependency is the greatest enemy of predictability. Intersections with cyclic control are easily predictable, even from the outside by observing its phases changes. With increasing traffic-dependency, predictability decreases until the incapacity of doing any prediction. With increasing traffic, traffic-dependency decreases due to oversaturation and predictability increases. Countdown devices installed at vehicle and pedestrian lights require good predictability of light changes, to red and to green. Higher predictability means lower traffic actuation and thus lower control quality and therefore longer travel times and longer waiting queues. Countdown devices have become very fashionable at many places. This paper tries to formalize the dependency between predictability and traffic dependency and tries to show ways how the two can cohabitate.

Abstract

Differences in traffic quality perception can easily be felt when traveling to different cities. At former World Congresses an approach for systematically measuring the quality of the control applied at individual intersections has been proposed under the main focus of pu...Lead moreDifferences in traffic quality perception can easily be felt when traveling to different cities. At former World Congresses an approach for systematically measuring the quality of the control applied at individual intersections has been proposed under the main focus of public transport priority quality [1]. A follow-up paper was focusing on detection possibilities for measuring or estimating loss time for individual traffic [2]. Now we would like to focus on pedestrians and bicycles without detection and the travelling quality to be expected. The paper introduces a method that bases on statistical sources that generate distributions for pedestrians and bicycles that enable quality evaluation with signal group change data only. The paper puts a special focus on pedestrians and bicycles crossing several lights in a row.

Abstract

This article provides a high-level overview of the development of highly automated driving systems and il-lustrates challenging situations and use cases, in some cases leading to fatal accidents. Accidents with auto-mated vehicles (AV’s) and Automated Guided vehicle (AGV’...Lead moreThis article provides a high-level overview of the development of highly automated driving systems and il-lustrates challenging situations and use cases, in some cases leading to fatal accidents. Accidents with auto-mated vehicles (AV’s) and Automated Guided vehicle (AGV’s) may give clues to why accidents happen and what’s needed to avoid similar accidents in the future. Empirical data from 13 years of human machine in-teraction between Automated Guided vehicle (AGV’s), employees, visitors and patients in a hospital shed light on recent accidents with Tesla’s autopilot. Results indicate self-driving cars need a better model, i.e. a true 3D model of their spatial dimensions and how it fits with safe travel. Automotive sensors may not be able to predict safe travel with current performance. Results indicate there will continue to be accidents with self-driving vehicles as long as the automated vehicle does not have a better sense of self like humans do, and that there will be an ongoing risk as long as changes to software and sensors are not made to account for the discrepancy. We outline the impact of these use cases on system design, key technologies, and their technical realization for a highly automated driving system.

Abstract

Autonomous parking vehicles offer highly desirable functionality, freeing occupants from tedious driving in enclosed car parks. Furthermore, the low speeds and structured environments offer ideal features for fully self-driving cars. We leverage communication between car ...Lead moreAutonomous parking vehicles offer highly desirable functionality, freeing occupants from tedious driving in enclosed car parks. Furthermore, the low speeds and structured environments offer ideal features for fully self-driving cars. We leverage communication between car and an intelligent car park to enable level 5 parking and enhance the function with additional services. The car park provides external localization, access control, slot reservation including charging spots and interfaces with end-user devices for send/call functionality of the vehicle. The results presented in this paper have been evaluated in a real-life car park using a self-driving vehicle.

Abstract

In road transport, efficient routing and navigation of vehicles is an essential means for saving time, fuel and emissions. Our vision is to address the problem of efficient routing by means of embedded, context-aware machine learning and advanced data processing technolog...Lead moreIn road transport, efficient routing and navigation of vehicles is an essential means for saving time, fuel and emissions. Our vision is to address the problem of efficient routing by means of embedded, context-aware machine learning and advanced data processing technologies. Our goal is to create compact, embedded automotive devices that will measure and keep history of fuel and time spent upon each road segment of the digital map, together with contextual parameters (environmental and vehicular conditions), so as to eventually learn from collected data, and predict consumption in future instances. This proposed technique will enable highly accurate and reliable routing costs estimation and, thus, routing optimisation. Nevertheless, it also poses great scientific and research challenges, as the learning and prediction process for an even small number of road segments can take hours to complete. Thus, the routing costs calculation time can become prohibitive. To this end, we will research and develop two families of solutions: hardware-based implementations of neural networks, and fast, embedded software-based implementations of learning models, with a view to achieving an execution speed orders of magnitude better than typical machine learning simulation software. Applications can be envisaged both for conventional vehicles and for automated driving scenarios.

Abstract

L3Pilot is the first comprehensive test of ADFs with hands-off the wheel on public roads across Europe. L3Pilot will test ADFs in 100 cars with 1,000 drivers across 10 different countries in Europe. The tested functions will be mainly of SAE automation level 3, some of th...Lead moreL3Pilot is the first comprehensive test of ADFs with hands-off the wheel on public roads across Europe. L3Pilot will test ADFs in 100 cars with 1,000 drivers across 10 different countries in Europe. The tested functions will be mainly of SAE automation level 3, some of them of level 4. This paper describes the data management toolchain we have designed and developed in order to exploit pilot data for answering a set of research questions about evaluation of such aspects as: technical and traffic, user acceptance, impact, socio-economic impact. The toolchain, supporting various confidentiality levels (prototype vehicle owner, consortium, public), has been designed to allow cross-vehicle owner data management, with efficient storage of data and its iterative processing with a variety of analysis and evaluation tools. Most of the tools in the data processing chain have been developed to a prototype version, tested in lab and are ready to be deployed for the pre-pilot starting in February 2019 in the various sites.

Abstract

Plug-in hybrid electric (PHE) buses combine advantages (range and flexibility) of buses with conventional internal combustion engines (ICE) with environmental and societal benefits (emissions- and noise-related) associated with electric buses. Thanks to geofencing and con...Lead morePlug-in hybrid electric (PHE) buses combine advantages (range and flexibility) of buses with conventional internal combustion engines (ICE) with environmental and societal benefits (emissions- and noise-related) associated with electric buses. Thanks to geofencing and connectivity, electric drive of PHE buses can be assigned to specific locations via a back-office system. Consequently, PHE buses not only can fulfil zero-emission (ZE) zones required by city authorities, but can also drive in electric mode in locations where the efficiency difference between electric motor and ICE is largest. Such a location-controlled behaviour allows executing targeted air quality improvement and noise reduction strategies as well reducing energy consumption. However, current ZE zone assignment strategies used by PHE buses are static—they are based on the first-come-first serve rule and do not consider traffic conditions. In this article, we demonstrate that dynamic assignment systems taking into account real-time parameters can significantly maximise environmental and efficiency benefits of PHE buses.

Abstract

Parking in town centres is a challenging problem that can be addressed from multiple lines of attacks. If informed solutions based on metered parking spots are evolving in some cities, traditional on-street parking generally makes up the majority of the parking capacity. ...Lead moreParking in town centres is a challenging problem that can be addressed from multiple lines of attacks. If informed solutions based on metered parking spots are evolving in some cities, traditional on-street parking generally makes up the majority of the parking capacity. Then, a probabilistic approach is the only suitable one as there is never any guarantee that a vacant spot will still be available within the time needed to reach it. In this paper, we first define a probabilistic model that can be used to simulate realistically the vacancy of parking spots among streets of a city. It includes the possibility to get short time information about street occupancy, along with the prediction of how this availability evolves on the long run. Based on it, we present a methodology to evaluate the performances of a route returned by a parking strategy in terms of total expected time to reach a goal (i.e searching and walking). Then, we use this methodology to evaluate the gain, in time, of our parking strategy, as well as its sensitivity to the uncertainty of the actual parking difficulty.

Abstract

The paper introduces a model-based optimisation procedure for design of signal synchronisation and bus priority strategies and speed advisory to car drivers. The traffic model simulates car traffic as platoons that travel along links and checks if they have to be joined o...Lead moreThe paper introduces a model-based optimisation procedure for design of signal synchronisation and bus priority strategies and speed advisory to car drivers. The traffic model simulates car traffic as platoons that travel along links and checks if they have to be joined or split at nodes where they are stopped. The model represents bus movements and stops along links individually and simulates the effect of different bus priority rules, applied online through bus identification devices. The design procedure is based on a hybrid genetic-hill climbing algorithm that optimises an objective function of delays for both drivers and transit passengers. Delays are computed by running the traffic model and simulating the online application of transit priority rules. The procedure has been applied in a real test case of a main street in Rome with tram lines on a reserved lane. Results obtained show that offline signal optimisation and online signal priority can significantly reduce both travel times of bus riders and delays for total traffic. Similarly, speed advisory to drivers, if considered in signal optimisation, can improve not only drivers’ delays but even transit passengers’ delays that travel on separated tram lanes, because it allows a better use of the road.

Abstract

The adoption of electric vehicles are an environmental friendly initiative and an economic opportunity since the number of these vehicles in the globe would be over 35 million by 2022. However, planning the routes of electric vehicles for timely delivery of goods, involve...Lead moreThe adoption of electric vehicles are an environmental friendly initiative and an economic opportunity since the number of these vehicles in the globe would be over 35 million by 2022. However, planning the routes of electric vehicles for timely delivery of goods, involves taking into account the limited range of the batteries, which may lead to detours for recharging purposes. In this work, the studied vehicles are modular. It means that they consist of an innovative system with one cabin module where the driver sits and one or more payload modules for the freight. The research problem arising when routing a fleet of such electric modular vehicles is relatively new. It extends the well-known Vehicle Routing Problem with Time Windows in the sense that its objective is to minimize the total routing costs while taking into consideration the time windows for serving the customers, the capacity of the vehicles but also the modular property of the vehicles and the recharging constraints. Therefore, an enhanced evolutionary method is proposed. More precisely, the evolutionary schema, which has the advantage to be population based, is hybridized with two local search strategies. An experimental study is conducted to compare the different approaches.

Abstract

This paper explores a concept to collect and utilise data from vehicle-installed sensors for the purpose of enhanced winter road management. Data from vehicle-installed sensors extends the coverage of road condition information from fixed Road Weather Stations. The paper ...Lead moreThis paper explores a concept to collect and utilise data from vehicle-installed sensors for the purpose of enhanced winter road management. Data from vehicle-installed sensors extends the coverage of road condition information from fixed Road Weather Stations. The paper introduces an open data platform - WiRMa - that is being piloted in the Finnish Lapland and analyses the prospective actor ecosystem around the platform. The platform’s data is aimed to be used for ‘information service packages’. A pre-assessment on the value of data is given with some evaluation on the distributional effects of the value. The prospective value and benefits gained from the data and information are allocated to potential users or beneficiaries. These users and beneficiaries, as well as the providers of the data and information, and the suppliers of technologies and services, form a business ecosystem of winter road management. The outcome of the analysis is a statement of benefits for the ecosystem. This statement is a justification and motivation for the platform investment. The actors need to commit to the platform to realise the potential gains. Some challenges and issues in achieving the commitments and a working ecosystem are discussed in the end of the paper.

Abstract

Nowadays technology and hardware reached the ability to run complex computational algorithms without costly requirements. Most of the modern large cities equip with sufficient powerful monitoring systems of traffic flow and surveillance cameras. It allows to expand the fi...Lead moreNowadays technology and hardware reached the ability to run complex computational algorithms without costly requirements. Most of the modern large cities equip with sufficient powerful monitoring systems of traffic flow and surveillance cameras. It allows to expand the field of intelligent transport systems (ITS) and create a safe city of the 21st century. ITS include many tasks, but most important of them is road video scene analysis and objects anomalies detection. Solving this problem will reduce the incidence response time, allow to predict further incidents and will help to optimize traffic flow. Tremendous progress in the field of computer vision and neural networks for object recognition task inspired community for active work in the field of image and video processing. Therefore, video cameras have become an integral part of the road monitoring systems. This work is dedicated to the actualization of problems and tasks of intelligent transport systems on the example of the city of Kazan. We will look through the problem of anomaly detection and propose possible methods for its solving based on modern machine learning techniques such as neural networks.

Abstract

Highly automated vehicles can fundamentally change road transport and improve life quality. Autonomous driving capabilities, provided by highly automated vehicles are enabled, facilitated and supported by the latest advances in Information and Communication Technologies (...Lead moreHighly automated vehicles can fundamentally change road transport and improve life quality. Autonomous driving capabilities, provided by highly automated vehicles are enabled, facilitated and supported by the latest advances in Information and Communication Technologies (ICTs), forming the cornerstone of the new generation of Intelligent Transport Systems (ITSs). The goal of this paper is to introduce a novel on-board intelligent management functionality that automatically and dynamically proposes the optimum autonomous level to the driver when he/she wants to make a certain journey with his/her automated vehicle. The proposed system utilizes (i) driver’s personal preferences, (ii) information based on sensor data fusion systems, and (iii) previous knowledge and experience on driving automation parameters. Knowledge is obtained through the exploitation of Bayesian networking principles and specifically the Naïve-based model. Indicative simulation results showcase the effectiveness of the proposed functionality, mostly with regards to proactively identifying an optimum autonomous level and accordingly notifying the driver.

Abstract

The occurrence of deer-vehicle collisions (DVC) is still an unsolved problem. There are many well-known different approaches with the aim of DVC reduction, but with no final solution. The effectiveness often can not be proven (light reflectors) or the measure has disadvan...Lead moreThe occurrence of deer-vehicle collisions (DVC) is still an unsolved problem. There are many well-known different approaches with the aim of DVC reduction, but with no final solution. The effectiveness often can not be proven (light reflectors) or the measure has disadvantages (fencing). The approach introduced in this paper contains a driver-side method, which was developed on the basis of 10 years accident data and linked additional data (weather data, calculation of light condition and environment data). The assessment of the risk is exemplarily carried out on accidents with roe deer, and is initially based on values of statistically determined categories. The risk index introduced in this paper can be linked with navigation systems and can be used for informing the driver of a current high risk because of time and spatial conditions. The approach can be further expanded with more variables and for other animal species.

Abstract

Nowadays spatio-temporal data analytics takes a big role for different applications with geographic information systems (GIS) including emergency services and Emergency Management Information Systems (EMIS). Using EMIS allows authorities to improve the management of emerg...Lead moreNowadays spatio-temporal data analytics takes a big role for different applications with geographic information systems (GIS) including emergency services and Emergency Management Information Systems (EMIS). Using EMIS allows authorities to improve the management of emergency events (fires, traffic accidents and traffic jams etc) in smart city. But to be EMIS more effective it must support spatio-temporal data analytic techniques aided to discovering new knowledge from the existing one, to events forecasting, detection the root events of the incidents etc. Moreover, due to a big volume of spatio-temporal data which needed to analysed, we must use relevant instruments for that. However the stack of modern technologies dealing with big data (Apache Hadoop, Spark etc) does not provide natively supported functions for spatio-temporal data analysis, so corresponding algorithms and libraries have to be developed for modern EMIS. In this paper we developed and evaluated some algorithms as well as a framework for big spatio-temporal data mining in Russian EMIS GLONASS+112. We suggested spatio-temporal co-location patterns mining technique and considered spatio-temporal outliers detection, spatial autocorrelation analysis algorithms. We evaluated the scalability and the time performance the algorithms.

Abstract

This research is an explorative study to look for the potential to predict traffic density from driver behaviour using signals collected from the Controller Area Network (CAN) bus. The hypothesis is that driver behaviour is influenced by traffic density in such a way that...Lead moreThis research is an explorative study to look for the potential to predict traffic density from driver behaviour using signals collected from the Controller Area Network (CAN) bus. The hypothesis is that driver behaviour is influenced by traffic density in such a way that an approximation of the traffic density can be determined from changes in the driver behaviour. The results are encouraging that the correlation between driver behaviour and traffic density can be established. An overall accuracy of over 95% is achieved with a precision of 92%. The recall rate however is low most likely caused by over-fitting due to the unbalanced dataset. The results still look promising and more training data should improve the results.

Abstract

AENA, the leading airport operator in terms of passenger volume worldwide, chose Sensefields intelligent traffic management technology to be installed on the airport of Barcelona in order to facilitate vehicle traffic control monitoring through an integrated automated sys...Lead moreAENA, the leading airport operator in terms of passenger volume worldwide, chose Sensefields intelligent traffic management technology to be installed on the airport of Barcelona in order to facilitate vehicle traffic control monitoring through an integrated automated system. Based on Sensefields magnetic sensor network that allows for advanced mobility data analytics solutions, the traffic flow on the entire airport premises is regulated depending on the real time conditions at both airport terminals. This paper presents the technology and solutions applied in order for the effective traffic vehicle control to become part of the supervisory control and data acquisition (SCADA) system of the busy Barcelona airport.