The Trainspotting Project

Detecting passing trains with miniature sensing units

In this project, we collected a data set of multiple days at several rail sections in Germany with a set of inertial sensing units. Each sensor node collected continuous raw acceleration data (at 100Hz) while being deployed at the rails. The data set shows that from the resulting vibrations, important information such as train types, speeds, directions, and wagon configurations can be estimated.
The aim of this data set is to evaluate combinations of easy-to-compute features that can be used on a sensor node for on-line classification of these 'train events'.

Publications:

  • Eugen Berlin and Kristof Van Laerhoven, "Trainspotting:
    Combining Fast Features to Enable Detection on Resource-constrained Sensing Devices". In INSS 2012, IEEE Press, 2012.
  • Eugen Berlin and Kristof Van Laerhoven, "Sensor Networks for Railway Monitoring: Detecting Trains from their Distributed Vibration Footprints", In DCOSS 2013, Cambridge, MA, USA, IEEE Press, pp. 80-87, 05/2013.
# # #

Datasets:

  • INSS'12 Dataset: 36 hours of data for a single sensor, 100Hz acceleration, ±250 passing trains
  • DCOSS'13 Dataset: 6 nodes placed along the tracks, 10 metres apart, 186 trains passed

Follow this link.

Log in