The ability to reliably predict public transport (PT) journey times is critical for network operators and transport authorities to measure, monitor and target improvements to the PT network, with flow-on effects for customers.
Research conducted in New Zealand between August 2016 and August 2017 aimed to identify and develop an optimal measure for PT predictability. This involved undertaking a local and international review of predictability/reliability measures used for PT or private vehicle travel, and included evaluation of measures.
From this review, and consideration of the potential for inter-modal and inter-regional aggregation, a shortlist of three preferred measures was developed including: buffer index, modified buffer index and planning index. Shortlisted measures were applied to a nationally aggregated set of PT travel data from across regions and PT modes.
This data testing helped assess ‘fit’ to the NZ Transport Agency’s road index, modification potential, and revealed that the shortlisted measures are all linearly related, with comparable results across different measures. This suggested there was not a compelling case for one particular measure to be used.
Validation workshops further revealed that stakeholders felt selection of any shortlisted measures depended on what aspect of reliability one wanted to examine and that care needed to be taken in comparing modes and developing thresholds.
Keywords: measures, performance, predictability, public transport, reliability