Project Website: ---


Start Date:
June of 2016

End Date:
November of 2018

@CRUiSE, An Advanced Tool for Cooperative Road Use

Road traffic agencies need to make informed decisions in order to define which road sections have the greatest risk of traffic-related impacts. In this context, the implementation of ATMS may not only improve network efficiency, but also minimize other traffic externalities. However, this implementation should consider both the level of contribution of each externality and its geographical scale; otherwise isolated measures may just migrate the problem elsewhere, there might be conflicts in minimizing different externalities. This can lead to undesirable side-effects and decrease the acceptability of ATMS. Also, despite the importance of the socio-economic impacts of transport, these large single indicators are still not easily perceptible by decision makers/citizens. Thus, in a context of increasing data availability, an important research topic is to unravel the complex relationships of these dynamic externalities, in order to efficiently manage the road network. Data integration is often a difficult problem, since data is usually structured in ways that do not easily match.

This research will be conducted within a formal partnership between the Transportation Technology Research Group of the Centre for Mechanical Technology and Automation (TEMA), the Research Group on Emissions, Modelling and Climate Change of the Centre for Environmental and Marine Studies (CESAM) and the Institute of Telecommunications (IT) at Aveiro, along with the Institute for Transportation Research and Education (ITRE) at North Carolina State University (NCSU), USA. The research team involves different research backgrounds, including transportation engineering, mechanical engineering, environmental engineering and computer science. The proposed collaboration between UA and NCSU represents a continuation of a long history of collaboration and achievements, such as joint publications (available here: http://transportes-tema.web.ua.pt/) and graduate students co-advising, and a joint Memorandum of Understanding to develop collaborative research in the field of Transportation Systems and related activities.

A more efficient management of existing infrastructures has been identified by EU as key strategy to reduce transport externalities. Road traffic agencies need to make informed decisions in order to define which road sections have the greatest risk of traffic-related impacts and hence prioritize the road sections in which an intervention is required. In this context, the implementation of ATMS could not only improve network efficiency, but also to minimize traffic externalities (traffic congestion, emissions, noise and safety). This can be done through smart traffic assignment strategies; (e.g. eco-routing, smart road pricing, variable message signals) or by better controlling the flow of vehicles in a certain road section (e.g. ramp metering, dynamic speed limits). However, the implementation of ATMS in the transport sector should consider both the level of contribution of each externality and its geographical scale; otherwise, some isolated measures may just move the problems elsewhere and perhaps causing greater damages to local communities. Thus, in a context of increasing data availability, a hot research topic is to unravel the complex relationships of these dynamic externalities, in order to manage efficiently current road networks.

This project will be a step forward compared with the state of art in different domains: from the concepts behind the impact assessment models to the optimization algorithms and the attention given to the user’s behaviour. Thus, the main research contributions of this project for the field will be: 1) impute Network Traffic Properties from High Resolution FCD and 2) inclusion of safety/energy/environmental parameters in the assessment of the road network performance, adjusted to local contexts of vulnerability.

A key feature of the @CRUiSE project is the recognition of the heterogeneity of the effects and their vibrant variation over time and space. Thus, it is intended that the decision maker can act in the transport system having a through understanding about the different impacts of the transport system and the main vulnerabilities associated with each link of the network. The main objective will not be to minimize itself a particular parameter, but rather to provide integrated solutions and holistic approaches capable of responding to the questions: What (to minimize)? When? Where? And how? This will be done by accurately assigning dynamic link-based indices of vulnerability and by providing an instrument for evaluating the cost and environmentally effectiveness of different ATMS and suggesting action plans to “where” and “when” they are most necessary.

Given the potential lack of accuracy of macroscopic tools and the difficulty in managing traffic in real-time combined with the use of microscopic models, link-based functional relationships between speed microscale patterns data of individual vehicles and real time macro scale traffic measurements will be developed in order to facilitate integration into optimization algorithms. Links with high intra-variability of speed, road grade and acceleration profiles will be analysed in detail.

The availability of geo-referenced data is increasing quickly, either from nomadic devices as well as from social media, and monitoring sensors networks. One of the challenges of this project is to combine and to increase the potential of each source of information by using innovative data streaming and data mining techniques. One of the main goals of @CRUiSE is to develop a common platform to gather real time information of different sources (see Figure 2) and use this platform to support optimization toolbox at centralized (planning level) and for providing (decentralized) information so that each network user can take advantage of safer roads, better environment and adopt more sustainable choices. Moreover, achieving an accurate prediction of road users’ acceptance to new eco-friendly suggestions is a very challenging task.

The first specific objective will be to develop a GIS-based dynamic map structure which will assimilate both historical data and Floating Car Data (FCD) integrated into a library of forecasting traffic models and associated traffic-related externalities (air pollution/climate, noise, and road conflicts). State-of-the-art instantaneous emissions and noise models (Cellular automota based approach models) will be integrated with existing traffic models. Consecutively, several air quality scenarios will be performed based on statistical models in order to shape a spatiotemporal database of pollutant concentration levels and identification of critical pollution hotspots as function of different congestion scenarios and weather conditions. Road conflicts and incidents will be analysed using safety models. A key feature of this platform is the recognition that these impacts are spatiotemporally dynamic due to the heterogeneity of activity patterns of each link as well as the air pollution levels/weather conditions. Accordingly, dynamic weights using an economic risk management approach for each impact will be assigned in order to develop a unique and dynamic link-based vulnerability index. In this work the concept of vulnerability is related with the potential for a population group (including drivers and users of a certain road segment) to experience damage in response to the influence of traffic-effects. At this stage, the link-based activity patterns (e.g. the exposed population living/working within a certain distance of the link) will be determined based on empirical observation and/or geostatistical data. In a further development, remote sensing, radiofrequency and activity–space analysis technologies could be integrated. Special attention will be given to the presence of vulnerable road users like cyclists.

The second objective is to enhance the potential of new sources of traffic data to improve the networks efficiency. This will be done creating new methods for managing different sources of real-time information to determine as accurately as possible the energetic and environmental network performance. Several vehicles (private vehicles, taxis and cyclists) will be equipped with multiple systems for monitoring the dynamic each vehicle (smartphone application, GPS data logger) circulating in the network under various scenarios of congestion. Simultaneously, road measurements of some macroscopic traffic parameters (road occupation, traffic flow) will be made at critical points of the network. This will allow matching FCD with traffic data to reconstruct the state of traffic on the road segment of interest. Complementarily to this, innovative functional relationships between microscale speed patterns based on individual vehicles (floating car data - FCD) and different levels of macroscopic traffic performance scenarios will be developed for different study areas. A key innovative factor will be the inclusion of energetic / environmental parameters in these relationships through the application of detailed state of-the-art traffic-related models.

Finally, the main deliverable of this project will be an integrated decision support system prototype for determining efficient traffic management measures. Based on diverse optimization algorithms and artificial intelligence techniques, a conceptual strategy to implement ATMS (e.g. optimal flow distributions, smart road pricing systems, optimum link speeds, eco-routing information) will be designed. A set of optimization toolboxes for traffic management will be developed in order to deliver an integrated optimization platform. Different solutions of network optimization will be assessed based on the perspective centralized traffic management (e.g. system-optimum traffic flow distribution and associated smart road pricing schemes) and under a decentralized perspective (user-optimum perspective, ex. optimal path/optimum speed) to ensure equitable and realistic solutions. This platform will be tested and demonstrated in two different cases-studies while it is intended that this methodology could be extrapolated to other regions.