The NSW Premier's Innovation Initiative is a significant opportunity for the NSW Government to engage with business, community and other non-government stakeholders to help us deliver services to the people of NSW. Under this initiative, Data61 has partnered with the NSW Government to compile data from Opal, GPS devices, traffic signals, and buses to provide better information for the Transport Management Centre to alert road users of congestion incidents and provide information about alternatives.
The objective is to develop a technical information service for Transport for NSW that provides information intended to augment, not replace, existing transport management practice. The service is provided by an instance of the Advanced Data Analytics in Transport (ADAIT) platform. ADAIT is a generic distributed computing platform developed by Data61 that is built from distributed databases, a parallel computing architecture, data analytics modules and defined interfaces.
The platform will integrate and fuse transport data across different modes from current data sources, and can be extended to add future data sources. The fused data is processed by the platform to provide situational awareness and predictive analytics of the transport network intended to inform road congestion management. The predictive analytics can account for changes in demand, behaviour, capacity, incidents or combinations of these. The system can also detect situations that deviate from normal and draw these to the attention of transport operators.
The service is provided through interactive visual interfaces and through APIs.
Situation awareness is the capability to monitor real-time operations of transport networks and identify operational anomalies in a way transport operators and travellers can use for making better decisions. This example demonstrates the platform capability to integrate SCATS geospatial data with real-time traffic detector information, including traffic volume, level of congestion, capacity utilisation and phase (green) time allocation.
Multi-modal analytics is the capability on using data from both road networks and public transport networks to conduct inference and prediction on:
- Multi-modal OD demand: predict the demand between origin and destinations in respect to all modes; such capability is current not available in the market.
- Multi-modal behaviour: predict the impact of service changes to modal, route, departure time choices for each identified user classes; such capability is currently not available in the market.
- Congestion factor analytics: identify the predominate factors for the anticipated congestion in service.
This example visualises the travel demand pattern at morning peak hours (7:45am) in Sydney, with Circular Quay as the final destination. The size of each cycle is proportional to its travel demand, with the cross centres on the location of the journey origin. The cycle in blue colour is for ferry, orange colour for train, and green for bus. This particular OD Analysis function allows user to playback the travel pattern of a chosen day, for any selected travel zone, either as origin or as destination.
Decision Support aims to use data analytics to derive additional insight and actionable outcomes for the transport agencies to promote the best practice in managing transport operations. This example demonstrates the capability of using GTFS data, a real-time public transport data feed openly available through TfNSW Open Data Hub, to derive insight on bus operations in Sydney. This function allows the user to query any chosen bus stop/route/zone in Sydney for its bus operations performance, measured by % of arriving early, on-time, late and by average minutes in delay and early. The user would be able to customise the configurations for the analysis, for example the threshold for the labelling on-time service.