Design, Modelling and Simulation

Integrated passenger behaviour, train operations diagnostics and vehicle condition monitoring system

Downer (Saad Khan, Mike Ayling, Loic Ayoul) / University of Technology Sydney (Alen Alempijevic, Stuart Warren)

Research summary

Together with Downer, the University of Technology Sydney created Dwell Track – a complex dwell-time diagnostics tool that uses 3D Infrared cameras and algorithms to anonymously monitor passenger numbers and movement at train station platforms. The technology can quantify passenger congestion on platforms and near train doors that can lead to extended dwell times. Data from Dwell Track could enable train operators to improve platform management procedures and communications to passengers.

Start/end date

1 July 2014 to 30 November 2019

Total contracted budget (including in-kind)

$3,816,545

Key achievements

• The Dwell Track technology was able to automatically detect the platform edge, train door status and position/orientation of up to 50 passengers per camera on a railway platform.
• 16 Ethernet connected depth cameras were installed along the length of Wynyard’s platform three in the Sydney Trains network in late 2019, providing real-time data on mobile devices of platform staff via dashboards in a trial of the technology.
• DwellTrack was launched by Downer and UTS at the CeBIT technology exhibition in Sydney in 2018.

Publications

Al-Widyan, F., Kirchner, N., and Zeibots, M., An empirically verified Passenger Route Selection Model based on the principle of least effort for monitoring and predicting passenger walking paths through congested rail station environments, Proceedings of the Australasian Transport Research Forum 2015.

Colborne-Veel, P., Kirchner, N., and Alempijevic, A. Towards more train paths through early passenger intention inference, Proceedings of the Australasian Transport Research Forum 2015.

Collart, J., Fitch, R. and Alempijevic, A. Motion States Inference through 3D Shoulder Gait Analysis and Hierarchical Hidden Markov Models. Australasian Conference on Robotics and Automation 2017, 2017, pp. 1 – 8

Collart, J., Kirchner, N., Alempijivic, A. and Zeibots, M. Foundation technology for development of an autonomous complex dwell time diagnostics tool, Proceedings of the Australasian Transport Research Forum 2015.

Farhood, H., He, X., Jia, W., Blumenstein, M. and Li, H. (2017). Counting People based on Linear, Weighted and Local Random Forests. Appeared at the Proceedings of 2017 International Conference on Digital Image Computing: Techniques and Applications (DICTA 2017).

Kasmani, S., He, X., Jia W., Wang, D. and Zeibots, M. (2018). A-CCNN: adaptive CCNN for density estimation and crowd counting, 2018 IEEE International Conference on Image Processing (ICIP2018).

Kirchner, N., Caraian, S., Colborne-Veel, P. and Zeibots, M. Influencing passenger egress to reduce congestion at rail stations. Proceedings of the Australasian Transport Research Forum 2015.

Virgona, A., Kirchner, N. and Alempijevic, A. Sensing and perception technology to enable real time monitoring of passenger movement behaviours through congested rail stations, Proceedings of the Australasian Transport Research Forum 2015.

Virgona, A., Alempijevic, A. and Vidal-Calleja, T. (2018). Socially Constrained Tracking in Crowded Environments Using Shoulder Pose Estimates. 2018 IEEE International Conference on Robotics and Automation (ICRA), Brisbane, QLD, 2018, pp. 1-9.