Towards the modelling of dynamic train systems: Disruption pattern analysis

Public transport such as train systems, particularly commuter line, become one of important transportation systems in Indonesia. However, there is still a challenge on how to ensure the reliability of such systems when respond to the changes or disruptions in operational time. This paper proposes a disruption pattern analysis of commuter line systems, as the first step towards building a dynamic model of train systems which can respond appropriately to those disruptions. The model will use a dynamic network as concurrent systems approach, which consists of two processes, namely: disrupter (a process that has a role to disrupt the system) and organizer (a process that has a role to normalise the system). We collect the disruption data of a commuter line system from its official account in the social media. The disruption pattern is analysed using Bayesian Network to obtain the relation between disruption parameters and their probabilities. The result of this analysis will be used in designing the automata of the disrupter.