Traffic congestion occurs daily, which can have negative effects on not only the quality of mobility, but also other important aspects of life like economic growth, health and environment. Both understanding and efficiently managing traffic are therefore crucially important tasks. Vast amounts of data are collected daily to gain insights into the dynamics of traffic. However, these data are typically stored in the form of raw measurements, that might hamper their potential benefits to both researchers and practitioners. A more informative and compact way to store traffic data is in the form of spatio-temporal maps, which have been shown to have advantage in intuitively observing traffic states. However, collecting, managing and retrieving such 2D patterns of congested traffic on large networks are challenging tasks. Accordingly, this dissertation is dedicated to developing methodologies and tools to advance the utilisation of traffic data, in particular, congestion patterns. A conceptual framework for an intelligent search engine for congestion patterns (socalled CoSI) is designed. It covers the entire requirements necessary to develop such a system, ranging from processing raw data to searching through the resulting database of congestion patterns. Overall, the framework consists of two parts: database construction and search application (or so-called pattern retrieval). Their designs and relations are comprehensively presented in this research. The database construction is responsible for preparing a database of patterns of congested traffic which is carefully designed for the conveniences of a search application. Its conceptual design consists of three layers (or phases): pattern collection, feature extraction and pattern annotation. Regarding the search application, several possibilities for retrieving patterns are identified in association with the aforementioned steps of constructing the underlying database. @en

Highway bottlenecks are responsible for the majority of traffic congestion. Although the problem of bottleneck detection is not new, contemporary methods have not solved the problem thoroughly with regards to bottleneck locations, activation time, and related congestion tracking. These elements are essential for identifying and characterizing a bottleneck. This paper proposes a comprehensive framework for detecting and extracting these features of highway bottlenecks from traffic data. We particularly focus on questions (i) whether a bottleneck is the primary source of congestion or (ii) whether it is activated due to congestion caused by another downstream bottleneck. The underlying principles of the proposed method include the detection of congestion (in spatio-temporal patterns of traffic congestion), and the detection of speed discontinuities in traffic data (since this is an important indicator of a bottleneck activation). The method is data-driven and automatic therefore can be easily applied to different highways and used to obtain meaningful statistics of existing bottlenecks. We have tested the method on simulated data and also demonstrated it on real data from a busy highway section in the Netherlands. The results suggest that the method is robust to different implementations, i.e. locations, of loop-detectors which measure traffic at discrete locations. @en

After developing the methodology, we applied it to the case of the Netherlands highways. We show that the average irradiation on the Dutch highway network is around 880 kWh/m2/y, 35% less than the potential of an optimally tilted conventional PV system in the south of the Netherlands. Covering the entire 1600 km of the Dutch highways network with solar road modules of poly c-Si, mono c-Si and CIGS would respectively generate 5.2 TWh/y, 6.6 TWh/y, and 3.4 TWh/y of DC electricity. This could be used to fully power the Dutch national public lighting demand. Moreover, to include the effect of traffic on these values, a model was developed to account for the energy potential reduction due to vehicles shading. Using real traffic data from two of the top-four busiest roads in the Netherlands, the A12 and A16, it was found that traffic accounts for an average of 3% reduction of solar road irradiation and DC yield potential. The maximum reduction of 9% was observed in particular locations, such as bridges and nearby ramp roads. The result of such mapping methodology could serve as a useful tool for research advisory, private industry, and governmental projects.@en

Classification of congestion patterns is important in many areas in traffic planning and management, ranging from policy appraisal, database design, to prediction and real-time control. One of the key constraints in applying machine learning techniques for classification is the availability of sufficient data (traffic patterns) with clear and undisputed labels, e.g. traffic pattern X or Y. The challenge is that labelling traffic patterns (e.g. combinations of congested and freely flow areas over time and space) is highly subjective. In our view this means that assessment of how well algorithms label the data should also include a qualitative component that focuses on what the found patterns really mean for traffic flow operations and applications. In this study, we investigate the application of clustering analysis to obtain labels automatically from the data, where we indeed first qualitatively assess how meaningful the found labels are, and subsequently test quantitatively how well the labels separate the resulting feature space. By transforming traffic measurements (speeds) into (colored) images, two different approaches are proposed to extract the features of a large number of traffic patterns for clustering: point-based and area-based. The point-based approach is widely applied in the image processing literature, and explores local interest points in images (i.e. where large changes occur in color intensity); whereas a new area-based approach combines domain knowledge with Watershed segmentation to partition the images into different spatial-temporal segments from which domain specific features, such as wide moving jam patterns, are extracted. The results show that the Watershed segmentation separates the traffic (congestion) patterns into more meaningful and separable classes, comparable to those that have been proposed in the literature. Since there is no ground-truth set of labels, the quantitative assessment tests how well both methods are able to separate the respective feature spaces they construct for the (large) database of traffic patterns. We argue that the more crisp this separation is; the better the labelling has turned out. For this quantitative comparison we train a multinomial classifier that maps unseen patterns to the labels discovered by each of the two labeling approaches. The most important result is that the classifier using the area-based feature vector achieves the highest average levels of confidence in its decisions to classify patterns, implying a highly separable feature vector space. We argue this is good news! Not only does the combination of image processing (Watershed) and domain knowledge (traffic flow characteristics) lead to meaningful labels that can be automatically retrieved from large databases of data; this method also leads to a more efficient separation of the resulting feature space. Our next endeavor is to further refine and use this method to develop a search engine for the (rapidly growing) 200 TB historical database of traffic data hosted by the Dutch National Datawarehouse (NDW).@en

Is it possible to use just aggregate carriageway data for the evaluation of congestion warning systems (CWS) in large networks—or any system affecting traffic safety for that matter? In this paper, two hypotheses related to this question are tested. The first hypothesis is that it can be done by comparing large-scale congestion patterns on road stretches with and without CWS. The underlying rationale is that heterogeneous congestion patterns with many disturbances, frequent wide moving jams, and large speed differences result in more potentially unsafe traffic conditions than more homogeneous congestion patterns. The second hypothesis is that it is possible to compare differences in average (maximum) deceleration distributions into congestion waves between road stretches with and without CWS. Both hypotheses have been tested for similar bottlenecks with similar demand patterns and the results suggest the first hypothesis must be rejected. Although the idea seems plausible (CWS result in more homogeneous congestion patterns) there were too many confounding factors in the data to make the case. However, persuasive evidence was found for the second hypothesis. Statistically significant differences were found between (maximum) deceleration distributions on road stretches with and without CWS that suggest CWS do—as expected—contribute positively to traffic safety. It thus seems to be possible to monitor safety effects using just average speeds. However, the method is limited to providing relative comparisons. Furthermore, to fully rule out the effects of unobserved factors, more evidence and validation with microscopic data are needed.@en

Traffic state estimation is an important task that has attracted a lot of research effort in recent decades. The main goal of traffic state estimation is to turn measured data, which is normally noisy and incomplete, into meaningful information for further investigation, either offline or online (e.g. traffic management and control).@en

Identifying and classifying traffic and congestion patterns are essential parts of modern traffic management underpinned by the emerging intelligent transport systems. This paper explores the potential of using a combination of image processing methods to identify and classify regions of congestion within spatiotemporal traffic (speed, flow) contour maps. The underlying idea is to use these regions as (archetype) shapes that in many combinations can make up a wide variety of larger-scale traffic patterns. In this paper, use of a so-called statistical shape model is proposed as a low-dimensional representation of the archetype shape, and an active shape model algorithm coupled with linear classification is developed to classify the patterns of interest. Application of the proposed method is demonstrated with a preliminary set of speed contour maps reconstructed from loop detector data in the Netherlands. The results show that the extended active shape model can be used as a multiclass classifier. In particular, 70% of the traffic patterns in the test data were correctly classified with use of only two archetype shapes and simple logistic classifiers. The results point to the importance of use of expert knowledge by means of (a priori) manual classification of the training examples. This work opens many research directions, including semiautomated searches through traffic databases, automatic detection, and classification of new traffic patterns.@en