Why did a vehicle stop? A methodology for detection and classification of stops in vehicle trajectories

ABSTRACT

Trajectory data mining is a lively research field in the domain of spatio-temporal data mining. Trajectory pattern mining comprises a set of specific pattern mining methods, which are applied as consecutive steps on a trajectory with the goal to extract and classify re-occurring spatio-temporal patterns. Despite the common nature and frequent usage of such methods by the GIScience community, a methodological approach is missing so far, especially when it comes to the use of machine learning-based classification methods. The current work closes this gap by proposing and evaluating a machine learning-based 3-steps trajectory data mining methodology using the detection and classification of stop points in vehicle trajectories as example. The work describes in detail the applied methodologies with respect to the three mining steps ‘stop detection’, ‘feature extraction’ and ‘classification in traffic-relevant and non-traffic-relevant stops’ and evaluates six machine learning-based classification algorithms using a real-world dataset of 15,498 vehicle trajectories with 5,899 detected stops (thereof 2,032 manually classified). Due to its exemplary nature, the presented methodology is suited to act as blueprint for similar trajectory data mining problems.

KEYWORDS:

• Trajectory data mining
• blueprint methodology
• vehicle trajectories
• stop detection
• stop classification
• machine learning

Data and Codes Availability Statement

The road data are openly available on Open Data Austria at https://www.data.gv.at/, reference number 3fefc838-791d-4dde-975b-a4131a54e7c5 whereas the vehicle trajectory data cannot be made publicly available due to their containing information that could compromise the privacy of human subjects. The codes for stop detection and classification together with a stepwise description and sample stop point data are available on OSF at https://doi.org/10.17605/OSF.IO/CQSR8.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1. http://open.gip.gv.at/ogd/gip_standardbeschreibung.pdf.

2. Implementations are: Bagging: v.4.2 adabag (Alfaro et al. 2013); Random Forest v.4.6.14 randomForest (Liaw and Wiener 2002); Boosting: AdaBoost.M1v.4.2 adabag (Alfaro et al. 2013); XGBoosing: XGBoost v.0.82.1 xgboost (Chen et al. 2019); KNN: v.1.3.1 kknn (Schliep and Hechenbichler 2016); ANN v.7.3.12 nnet (Venables and Ripley 2002).

3. https://www.data.gv.at/katalog/dataset/intermodales-verkehrsreferenzsystem-osterreich-gip-at-beta; licensed as CC BY 4.0.

Additional information

Funding

This work was supported by the Austrian Federal Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology under Grant [GZ BMVIT-612.014/0008-III/I1-2015].

Notes on contributors

Karl Rehrl

Karl Rehrl holds a diploma degree in computer science from the University of Linz / Austria and a doctoral degree in geo-information from the Technical University of Vienna / Austria. He is heading the Mobility and Transport Analytics (MTA) research group at Salzburg Research, an applied research institute specialized in the field of Motion Data Intelligence (MDI). His research interests are in analysing and interpreting motion data in the field of mobility & transport, with an emphasis on Probe Vehicle Data (PVD) and Cooperative-ITS (C-ITS). Karl Rehrl has 15+ years of experience in initiating and heading applied research projects and pilot demonstrations and published 60+ scientific articles. He is an editorial board member of the Journal of Location Based Services.

Simon Gröchenig

Simon Gröchenig holds a master degree in Spatial Information Management from the Carinthia University of Applied Science / Austria. He is a member of the Mobility and Transport Analytics (MTA) research group at Salzburg Research, an applied research institute specialized in the field of Motion Data Intelligence (MDI). His research interests include the acquisition, analysis and visualization of traffic related spatial data, in particular Probe Vehicle Data (PVD) and road network data. Recent research also deals with data from autonomous vehicles and high-definition maps.

Stefan Kranzinger

Stefan Kranzinger completed his studies at the University of Linz / Austria with a diploma in business administration and economics and a master's degree in social economy.  Moreover, he wrote his dissertation at the Vienna University of Economics and Business / Austria and holds a PhD in Economics. He works at the Mobility and Transport Analytics (MTA) research group at Salzburg Research, an applied research institute specialized in the field of Motion Data Intelligence (MDI). His research is focused on the empirical analysis of mobility & transport data.