Prediction of disruption in public transportation
Discover more about this AI work within public transportation disruption management. Here, the exploration revolves around the ability to predict primary disruptions and real-time deviations from schedules to establish a more efficient and reliable public transportation system. Explore project outcomes, including project documentation and the latest technology for primary disruption prediction.
Result Achievements
Primary Disruption: After exploring the latest technology and based on the project team's assessment, it seems possible to predict primary disruptions. However, the increased probability of predicting a specific disruption is often relatively small. The expectation is that the use of these predictions will mainly apply to specific traffic information and for detecting common causes of disruptions, rather than being used for daily traffic planning and disruption management. Identifying common causes of disruptions can, however, also be done using traditional methods.
Secondary Disruption: After considering the costs of testing this technology, the project team believes that a full-scale test of usage scenarios would require collaboration and resources from the Swedish Transport Administration (Trafikverket) and several other authorities.
Deliverables include documentation of the project and a compilation of the latest technology for predicting primary disruptions.
What was the initial goal of the project?
The project team has explored two main disruption types within public transportation. The primary disruption focuses on the possibility of predicting deviations from the plan under normal traffic conditions before any disruptions occur. The goal was to create meaningful predictions that could be used for potential actions.
The secondary disruption investigated was the technology for real-time prediction of deviations from the timetable when a vehicle deviates. The project explored how this technology works and reviewed the possibility of identifying the prioritized usage areas within contracted public transportation. By analyzing and understanding these disruption types, they aimed to improve the efficiency and reliability of the public transportation system.
Background
Currently, most disruption management occurs reactively, manually, and often with difficulties in effectively communicating relevant information to passengers.
Watch the presentation of the project from the final conference
Partners
Partners in the project are Malmö Universitet, The Train Brain, RISE and Västtrafik.