An optimisation-based digital twin for automated operation of rail level crossings

Abstract

Railway level crossings (LCs), as the intersection of road and rail transport, are the weak points in terms of safety, as they are used by different modes of transport. The safety level at LCs can therefore be affected by the behaviour of the users. However, the level of safety can also be affected by failures and errors in the operation of LC equipment. Apart from safety, errors and failures of the LC devices can lead to longer waiting times for road users. As the volume of traffic on rail and road increases, so does the risk that the level of safety will decrease. The increase in traffic volume via LC leads to higher traffic volume on the road and more frequent trains on the rail, which leads to longer waiting times for road users on the LCs. The longer waiting times can disrupt the traffic flow, especially during peak hours when the growing volume of traffic on road and rail increases road user dissatisfaction. Moreover, in the era of Industry 4.0 and Digital Rail, new digital and automated technologies are being introduced to improve rail performance and competitiveness. These technologies are aligned with the LCs and are intended to ensure the efficient operation of LC and the efficient use of LCs by conventional trains as well. To achieve this, a concept is needed that simultaneously monitors and visualises the operation of LC in real time, identifies potential faults and failures of the LC equipment, and updates and monitors the proper operation of LC based on the historical data and information of the operation of LC according to the road traffic volume and the characteristics of the rail traffic and trains. Therefore, in this study, a digital twin system (DT) for rail LC was initiated and built as a concept that can meet the above requirements for proper LC operation in real time. DT of LC includes all components of LC and communication between them to synchronise the operation of LC according to the real-time requirements. The DT system is able to optimise the operation time of LC by monitoring the operation of LC and collecting data to ensure efficient use of LC and reduce unnecessary waiting time for road users. In this paper, the operation time of LCs on Swedish and Taiwanese railways was compared using the developed level crossing optimisation model (OLC). Since the introduction of new signalling concepts requires an improvement of LC operating characteristics and their design, the operating strategies were modelled using the OLC model. The results of the work show that the optimal values of LC operation time are different for the case studies investigated. The replacement of track circuits as detection devices and the introduction of balises can also positively influence the operation time, as well as increasing the speed of trains via LCs. However, due to the formulation of the OLC model, the impact of a longer train length on the operation of LC is not recognised. The OLC model can be used to estimate the real-time operation time of LC under different traffic conditions as well as the impact of different changes and extensions of LC.