In this study researchers developed a decision support tool to improve the performance of transit operations in Auckland, New Zealand. Using an agent-based simulation framework a model was constructed to represent bus operations across the network and generate random input data needed to emulate real-life traffic conditions where response plans can be implemented to limit impacts of schedule deviations on overall passenger travel times and wait times between transfers.
The analysis examined performance with and without three real-time operational control strategies (holding, skip-stops, and short-turning). Combinations of these strategies were modeled across fourteen scenarios to cover a wide range of potential schedule deviations as needed to identify optimal combinations of control tactics.
Implementing transit operational strategies that adjust to prevailing traffic conditions resulted in a 4.7 percent reduction in overall travel time for transit passengers and a 150 percent increase in the number of direct transfers (no waiting). Best performance occurred with strategies that used short-headways.