Motivated by the requirement to understand and manage the city-scale traffic system in Sydney, the ADAIT team since 2015 is developing a large-scale traffic simulation platform to assist transport planners and traffic engineers to evaluate the holistic impact of their proposals or solutions to the entire urban traffic system in Sydney.
The key is to leverage machine learning technologies to solve problems in a city-scale traffic simulation platform in Sydney, in order to make such a platform efficient and effective to deploy, execute and maintain . NICTA developed its own all-in-one traffic simulation model calibration package. The package establishes interfaces between the underlying traffic model and the ubiquitous traffic data. This allows automatic calibration of model parameters, in order to match simulated results with real-world surveillance measurements. The calibration package reduces the life cycle of a traffic simulation project by months.
A calibrated and validated city-scale traffic system model in Sydney produces insights of the traffic system, which are extremely valuable for transport planning and congestion management. Examples include:
- The platform analyses the connectivity between critical regions (e.g. Parramatta and the CBD) and outputs the level of demand, the key used routes and the gap between the demand and the supply.
- For recurrent or non-recurrent incident, the platform estimates the portion of affected trips, their origins, their destinations and possible detour routes.
- For the most congested roads, the platform identifies the portion of trips that contribute the congestion and corresponding congestion solutions.