Hannes Koller

Novel road classifications for large scale traffic networks

Establishing a highly sophisticated large-scale Traffic Information System (TIS) requires the creation and deployment of link travel time prediction models for large road networks. Due to the dimension of typical road networks and low coverage with Floating Cars (FC), data sets that can be used for prediction contain a large number of missing observations. Additionally, specifying prediction models for each link separately is impossible due to restrictions of both computational as well as modeling resources. This paper aims to improve the scalability of link travel time predictions by combining information from roads with similar characteristics. The Functional Road Class (FRC) is a widely accepted indicator for road similarity mainly based on static information from infrastructure planning. The coherence between the clustering introduced by the FRC and road dynamics measured by Floating Car Data (FCD) in the city of Vienna is discussed and analyzed. Clustering approaches that are based on indices characterizing speed measurement distributions are proposed as alternatives to the FRC system. It is demonstrated by way of examples that the new clustering is much more appropriate to provide predictions of link travel times.

Predicting Motorway Traffic Performance by Data Fusion of Local Sensor Data and Electronic Toll Collection Data

In this article data fusion from different sources in order to improve estimation and prediction accuracy of traffic states on motorways is proposed. This is demonstrated in two case studies on an intraurban and an interurban motorway section in Austria. Data fusion combines local detector data and speed data from the Electronic Toll Collection (ETC) system for heavy goods vehicles (HGV). A macroscopic model for open motorway sections has been used to estimate passenger car and HGV density, applying a standard state-space model and a linear Kalman filter. The resulting historical database of 4 months of speed-density patterns has been used as a basis for pattern recognition. A nonparametric kernel predictor with memory length of 9 and 18 hours has been used to predict HGV speed for a prediction horizon of 15 minutes to 2 hours. The results showed a good overall prediction accuracy. Correlation analysis analysis showed little bias of predicted speed for free flow and congested time intervals, whereas transition states between free flow and congestion were frequently biases. The prediction accuracy can be improved by applying a combination of different prediction methods. However, the computational performance of the predictor needs to be further improved prior to implementation into a traffic management center.

Robust road link speed estimates for sparse or missing probe vehicle data

Probe vehicles equipped with GPS can be used to permanently collect traffic speed information for an entire road network, and the statistical mean value of link speeds collected over time is often used as an estimator for mid-term predictions. For road links with sparse probe vehicle data, the estimated mean may be too inaccurate due to the low sample size, and speeds for road links with missing probe vehicle data must be imputed from other data. This paper proposes to apply a Gaussian-mixture based technique to increase the robustness of speed estimates. Typical shapes of the diurnal speed curve are learnt from historical data of all links in the road network. The model is able to provide robust estimates of mean speed curves based on only a few available observations and drastically reduces the amount of data needed to store them by 95%. Experimental results on a comprehensive set of 857527 day speed curves show that the predictions are superior to traditional approaches based on aggregated or disaggregated historical mean values.

Fast Hidden Markov Model Map-Matching for Sparse and Noisy Trajectories

The problem of map-matching sparse and noisy GPS trajectories to road networks has gained increasing importance in recent years. A common state-of-the-art solution to this problem relies on a Hidden Markov Model (HMM) to identify the most plausible road sequence for a given trajectory. While this approach has been shown to work well on sparse and noisy data, the algorithm has a high computational complexity and becomes slow when working with large trajectories and extended search radii. We propose an optimization to the original approach which significantly reduces the number of state transitions that need to be evaluated in order to identify the correct solution. In experiments with publicly available benchmark data, the proposed optimization yields nearly identical map-matching results as the original algorithm, but reduces the algorithm runtime by up to 45%. We demonstrate that the effects of our optimization become more pronounced when dealing with larger problem spaces and indicate how our approach can be combined with other recent optimizations to further reduce the overall algorithm runtime.

Estimating travel times from static map attributes

Estimating accurate travel times on road networks is a prerequisite for many mobility related applications such as transportation planning, dynamic intermodal routing or logistics. Two widely used methods are (1) the trivial but inaccurate calculation of travel times using static speeds taken from road maps and (2) the use of historic time series calculated from rich data sets. For the second method extensive measurement campaigns are required. In this paper we present a novel approach to estimate realistic travel times exclusively from static map coefficients without the need for further data collection. Our method uses a linear regression model to estimate the diurnal variation of travel times for cars in urban and interurban areas. We discuss the model which has been developed and calibrated for the city of Vienna, Austria, and demonstrate the transferability of the model to a different city.

Identifying faulty traffic detectors with Floating Car Data

Virtually all ITS applications rely on accurate traffic data. Identification of faulty detectors is thus vital for their reliability and efficiency. Most existing approaches solely use current and historical data of single or adjacent detectors and are based on empirical thresholds. We present a method for fault detection using Floating-Car Data (FCD) as independent source of information which allows to distinguish changed traffic conditions from sensor faults. Fault detection is based on residuals of a nonlinear regression model fitted to detector readings and FCD traffic speeds. Instead of applying rule-of-thumb thresholds we employ a statistical test, where thresholds result naturally from historical data, sample sizes and required fault detection accuracy. We provide a theoretical framework for fault detectability analysis and empirically evaluate the fault detection capability of our approach using data obtained from a microscopic traffic simulation.

Utilizing mobility data to facilitate the introduction of E-Taxis in Vienna: Feasibility study of a decision support system for the introduction of battery electric vehicles as Taxis

Introducing E-Taxi fleets in urban areas poses a number of economic, organizational and technical challenges related to the nature of Battery Electric Vehicles (BEV). This paper discusses these challenges and demonstrates how existing mobility data can aid the underlying decision process to overcome them. We present an integrated approach developed for the introduction of an E-Taxi system in the city of Vienna, where mobility data based on taxi floating car data (FCD) was used as decision support.

A data-driven approach for travel time prediction on motorway sections

In the presented case study, travel times for passenger cars (PC) and heavy goods vehicles (HGV) were predicted with a data-driven, hybrid approach, using historical traffic data of the entire high-ranking Austrian road network. In case flow data were available, travel time was predicted with a Kernel predictor searching for similar speed-density patterns. In case of missing flow data, travel time was predicted with deviations from typical historical speed time series. The performed steps in pre-processing traffic data, the hybrid prediction method as well as the results for selected road sections are described and analysed.

Synchronizing Trucks and Trailers in a Multiresource Problem with Variable Neighborhood Search

A real-world container drayage problem in which containers are transported between an intermodal terminal, a container terminal, and customer locations is considered. The problem was modeled as a multi-resource routing problem (MRRP) that included trucks, trailers, and containers. Given a fleet of trucks and trailers, the goal is to use these resources most efficiently to complete a number of given orders. Orders consisted of several tasks with time windows, such as picking up a container at the terminal, delivering it to a customer, and bringing the processed container back. A challenging aspect of this problem is the management of trailers, which are required to transport the containers. Here, the compatibility between container types and trailer types must be considered. Thus, the decision of which trailer should be attached to which truck depends on the containers that must be transported, the day of availability of trailers, and the toll costs of the truck and trailer combination on the highways. This paper presents an efficient way to model this problem and proposes a metaheuristic approach based on a variable neighborhood search. It uses a compact solution representation and tailored neighborhood structures to reduce the search space. Classical MRRP neighborhood structures, as well as problem-specific ones, were used in combination and contributed to the overall success. The results show that the given real-world problem can be solved efficiently, and it can be shown that with proper planning, the utilization of resources can be increased.

Improving vehicle speed estimates using street network centrality

ABSTRACT This paper describes a novel approach to improve prediction models which estimate vehicle speeds and their diurnal variation for road network links in urban street networks using only static map attributes. The presented approach takes into account previously neglected spatial information by integrating network centrality measures for closeness (indicating how central a link is) and betweenness (indicating how important a road link is) into the prediction model. The model is calibrated with a real-world dataset of 100 million individual speed measurements from a fleet of 3500 taxi probe vehicles in Vienna, Austria. Given that centrality can be derived directly from readily available street network data, the experimental results demonstrate that integrating centrality measures considerably improves the predictions without the need for adding a supplementary data source. Improvements for vehicle speed estimates are particularly prevalent on important street network links in the city center as well as rural streets in the periphery.