Search

This present document is Deliverable D3.2 entitled 'Cooperative manoeuvring in the presence of hierarchical traffic management', which was prepared in the context of the WP3 framework of the TransAID project. The scope of this document encompasses the modelling and simulation of cooperative manoeuvring in the context of the microscopic traffic simulation activities conducted within TransAID. Initially, the state of the art in the domain of cooperative manoeuvring is provided and then two different cooperative manoeuvring frameworks are introduced. The first one is a decentralized framework where cooperative manoeuvring is solely based on vehicle-to-vehicle (V2V) communications, while the second one is a centralized framework that utilizes vehicle-toanything (V2X) communications. The logic for simulating the decentralized approach in the microscopic traffic simulator SUMO is subsequently introduced along with the corresponding functionalities that were developed within SUMO for this purpose. Cooperative manoeuvring is coupled with hierarchical traffic management by explaining how the decentralized approach can be integrated in the traffic management plans that were developed for each use case examined in the context of TransAID. Cooperative manoeuvring is coupled with traffic separation in SUMO and a timeline of cooperative manoeuvring actions in the simulation is presented. Coupling with communications is also addressed. Moreover, adaptations to the vehicle/driver models, that were developed to replicate the behaviour of cooperative and automated vehicles (CAV), are proposed based on the findings of the real-world prototype experiments. Finally, focus on the centralized approach in terms of development of relevant SUMO functionalities, and integration within the TransAID traffic management plans will be placed during the second project iteration.

Aerial images potentially contain a wealth of information relevant to the prediction of road safety if they could be thoroughly analyzed in great numbers. Coincident with the widespread availability of satellite and aerial images, machine learning algorithms for image processing and automatic object detection and classification are maturing. This allows the automated processing of huge amounts of image data by artificial neural networks (ANNs) or related machine learning systems - an area in which convolutional neural networks have shown a significant improvement over conventional methods. In the submitted work initial results on the application of machine learning on aerial images are presented. The goal is to determine an estimation of road safety levels. ANNs were trained to predict crash frequencies for road intersections relying merely on aerial images of the intersections. The used data consists of police recorded crashes in the city of Berlin and aerial images provided by the Berlin Senate Department for Urban Development. The performance of the ANN suggests that the line of research is worth further pursuit. For instance, the trained ANN was able to predict the presence of crashes on intersections in a Berlin district excluded from the training process with an accuracy of approximately 74%.

Estimating road safety is a major concern of a large body of theoretical research as well as for practitioners all over the world. Most related studies rely heavily on structured data as tables concerning the road geometry, infrastructural items, traffic volumes, etc., which are not always available. A more and more universally available source of data, which has rarely been used in conjunction with road safety research are aerial or satellite images. These images potentially contain a wealth of information relevant to the prediction of road safety if they could be thoroughly analyzed in great numbers. Coincident with the widespread availability of satellite and aerial images, machine learning algorithms for image processing and automatic object detection and classification are maturing. This allows the automated processing of huge amounts of image data by artificial neural networks (ANNs) or related machine learning systems, an area in which convolutional neural networks have shown a significant improvement over conventional methods. In the submitted work initial results on the application of machine learning on aerial images are presented. The goal is to determine an estimation of road safety levels. ANNs were trained to predict crash frequencies for road intersections relying merely on aerial images of the intersections. The used data consists of police recorded crashes in the city of Berlin and aerial images provided by the Berlin Senate Department for Urban Development. The performance of the ANN suggests that the line of research is worth further pursuit. For instance, the trained ANN was able to predict the presence of crashes on intersections in a Berlin district excluded from the training process with an accuracy of approximately 74%.