UNSW Faculty of Science Industry Research Forum – RTA presentation Presented 6 October 2010 By Christian Chong-White Traffic Algorithm Development Manager Traffic Management Roads and Traffic Authority of NSW Contributed content sourced from Fraser Johnson, RTA and David Shteinman, MASCOS. Roads and Traffic Authority of New South Wales The Roads and Traffic Authority of New South Wales (RTA) is here to present: • background on its traffic systems capability • an industry research project • some personal thoughts about collaborative research. Traffic Systems The RTA (and its former constituent organisations) has a long history of traffic systems development. The Sydney Coordinated Adaptive Traffic System (SCATS) is an adaptive traffic control system that was first developed in the 1970’s. The RTA’s Traffic Management Branch continues to develop SCATS. Additional traffic systems products are available that complement SCATS. SCATS worldwide SCATS has been distributed to 141 cities in 24 countries, controlling almost 32,000 intersections worldwide. SCATS installations can be found in: •Australia •New Zealand •China •Singapore •Iran •Ireland •Poland •Columbia •Brazil •Mexico •USA •… SCATS on the web The SCATS web site can be found on the internet at: www.scats.com.au Network Performance Development Section Network Performance Development (NPD) is a section of Traffic Management Branch Fraser Johnson Manager, Network Performance Development Christian Chong-White Traffic Algorithm Development Manager Network Performance Development Section Network Performance Development (NPD) specialises in: • Traffic control policy – rules for traffic control systems • Traffic performance – travel time/delay, capacity and queuing • Traffic control system design – traffic signal control, ramp metering, lane control, motorway management systems, bus priority • Traffic behaviour related to traffic control systems • Traffic measurement from and for traffic control systems • Traffic analysis of traffic control system results • Traffic modelling RTA project. Business need. Traffic modelling plays an important part in advising decision makers on the impacts of traffic performance. As a result, a research topic in traffic modelling was identified… RTA project. Project title: “Development of a statistical framework to guide traffic micro-simulation studies.” Project partners: RTA (funder/mgr) – business problem and domain expertise. MASCOS – statistics and cross-industry expertise. Azalient – modelling technique and practice expertise. (MASCOS: Centre of Excellence for Mathematics and Statistics of Complex Systems) RTA project. People. Project People: David Shteinman* Sandy Clarke Gary Millar MASCOS / UNSW MASCOS / UMelb Azalient Fraser Johnson RTA Christian Chong-White** RTA (* key contact | ** project initiator / manager) RTA project. High-level objective. Project objective: The project aims to provide enhanced statistical rigour in the design, planning and evaluation of traffic simulation studies to provide measured confidence in results. RTA project. Outcome. Science-based guidelines for: 1. Statistically rigorous & effective analysis of model results. 2. Quantify confidence in simulation outputs. 3. Quantify & correct bias due to model scope (boundary conditions). 4. Diagnose model problems before finalising results. 5. Predict cost of simulation as function of network features, precision. This advice is not readily available to traffic modellers and stakeholders. RTA project. Guidelines. Part 1. Design Stage – Quantify precision required – Predict simulation costs as a function of precision and network complexity Part 2. Output analysis A statistical framework to – Automate/simplify analysis stage – Pair-wise comparison of model runs (Scenario A vs. B) – Diagnostic tests for simulation outliers – Method to identify critical interactions between variables (Source: David Shteinman 2010) RTA project. Expected contributions. • Quality example of productive investment in “business intellectual infrastructure”. • Outcomes can make a valuable contribution to RTA modelling practice in line with project objectives. • National advice through Austroads guidelines and standards. • Expected to result in design changes to traffic simulation applications. • International advice – academic paper and special session at Intelligent Transport Systems World Congress in Busan, Korea. • Potentially more publications and specifications in the future. RTA project. Debrief. • Project has proved a particularly successful example of a complex systems development project. • Initial project management design a key contribution: – crystallised vision of desired target deliverables. – tight scope, defined objectives and focussed methodology. • Collaboration between parties has worked very well: – parties formed clear roles and contributions. • Potential to re-apply the successful strategy both within the transport modelling problem and beyond. RTA project. Take away. The project demonstrates the opportunity to deploy specialised knowledge (University) to specific industry problems (industry). Industry often does not have the critical expertise. This is particularly true where industry is managing and outsourcing. This is particularly needed for the development of new and enhanced: policies, products and services. Concluding comments. Industry has many, many problems that need to be solved. The key challenges are: • identifying the problems that have the potential to maximise value. • identifying appropriate collaborating teams to develop productive solutions. • creating the initiative and the ‘matching’ to facilitate the exercise. The end. Thankyou.