Floating car data (FCD) denotes the type of data (location, speed, and destination) produced and broadcasted periodically by running vehicles. Increasingly, intelligent transportation systems take advantage of such data for prediction purposes as input to road and transit control and to discover useful mobility patterns with applications to transport service design and planning, to name just a few applications. However, there are considerable quality issues that affect the usefulness and efficacy of FCD in these many applications. In this paper, we propose a methodology to compute such quality indicators automatically for large FCD sets. It leverages on a set of statistical indicators (named Yuki-san) covering multiple dimensions of FCD such as spatio-temporal coverage, accuracy, and reliability. As such, the Yuki-san indicators provide a quick and intuitive means to assess the potential ``value'' and ``veracity'' characteristics of the data. Experimental results with two mobility-related data mining and supervised learning tasks on the basis of two real-world FCD sources show that the Yuki-san indicators are indeed consistent with how well the applications perform using the data. With a wider variety of FCD (e.g., from navigation systems and CAN buses) becoming available, further research and validation into the dimensions covered and the efficacy of the Yuki-San indicators is needed.

@en

Recently, cutting edge technologies to facilitate data collection have emerged on a large scale. One of the most prominent is the in-vehicle data recorder (IVDR). There are multiple ways to assign the IVDR's data to the different drivers who share the same vehicle. Irrespective of the level of sophistication, all of these technologies still suffer considerable limitations in their accuracy. The purpose of this paper is to propose a methodology, which can identify the driver of a given trip using historical trip-based data. To do so, an advanced machine learning pipeline is proposed. The main goal is to take advantage of highly available data - such as driver-labeled floating car data collected by a IVDR - to build a pattern-based algorithm able to identify the trip's driver category when its true identity is unknown. This stepwise process includes feature generation/selection, multiple heterogeneous explanatory models, and an ensemble approach (i.e., stacked generalization) to reduce their generalization error. Our goal is to provide an inexpensive alternative to existing driver identification technologies, which can serve as their complement and/or validation purposes. Experiments conducted over a real-world case study from Israel uncover the potential of this idea: it obtained an accuracy of 88% and Cohen's Kappa agreement score of 74%.

@en

Long-term travel time predictions are crucial for tactical and operational public transport planning in schedule design and resource allocation tasks. Similarly to any regression task, its success considerably depend on an adequate feature selection framework. In this paper, we approach the myopia of the State-of-the-Art method RReliefF on mining relevant inter-relationships of the feature space relevant for reducing the entropy around the target variable on regression tasks. A comparative study was conducted using baseline regression methods and LASSO as a valid alternative to RReliefF. Experimental results obtained on a real-world case study uncovered the bias/variance reduction obtained by each approach, pointing out promising ideas on this research line.@en

The rapid increase in automated data collection in the public transport industry facilitates the adjustment of operational planning and real-time operations based on the prevailing traffic and demand conditions. In contrast to automated passenger counts systems, automated vehicle location (AVL) data are often available for the entire public transport fleet for monitoring purposes. However, the potential value of AVL data in estimating passenger volumes has been overlooked. This study examined whether AVL data could be used as a stand-alone source for estimating onboard bus loads. The modeling approach infers maximum passenger load stop from the timetable and then constructs the load profile by reverse engineering through a local constrained regression of dwell times as a function of passengers flows. To test and demonstrate the potential value of the proposed method, a proof of concept was performed by conducting unsupervised experiments on 1 month of AVL data collected from two bus lines in Dublin, Ireland. The results suggest that this method can potentially estimate passenger loads in real time in the absence of their direct measurement and can easily be introduced by public transport operators.@en