A researcher from RMIT University has used simulation software to optimise railway crossings and increase traffic flow, which he claims could result in billions of dollars in savings related to upgrading infrastructure.The Departure Side Platform (DSP), which was built using VisSim, places the city-bound or arrival platform on the other side of the level crossing, where the arrival and departure platforms are diagonal to each other instead of face to face.This can reduce road traffic congestion at railway crossings by 50-70 per cent, according to the researcher who built DSP, Dr William Guzman.Guzman claims the DSP could save billions of dollars in infrastructure costs associated with upgrading railway station level crossings or building underground tunnels to alleviate congestion.The Victorian government is trying to lease the Port of Melbourne to raise money for various projects including upgrading 50 problematic railway crossings. Guzman has calculated that the $160 million on average it costs for a 'full-blown grade separation' could be reduced to $1 million per crossing with DPS.\u201cThere are 172 railway crossings in the Melbourne metropolitan area. And this problem occurs all around the world, not only in Victoria," he said.\u201cBy moving the platform, the boom gates are closed for less time than they are now and so by having less time closed, there is more traffic flow.\u201dGuzman used data from the Victorian transport department and VicRoads, including data on intersections, traffic volumes, and train timetables across the Melbourne network.He also collected times and closures of the station level crossings.\u201cThe timetables are not always exact because sometimes they have cancellations, so they get delayed. So I collected the data by staying at the platform for five hours at a time for 20 days. I stayed at the station with a stopwatch collecting data, and the road signals,\u201d he said.When starting his research journey seven years ago, Guzman knew the time differences between dropping his daughter off to school and driving back home didn\u2019t seem right. For many years he drove his daughter to school, stopping at a railway crossing along the way. It was over this time he started to figure out a problem.Guzman said he would often wait up to 10 minutes at the railway crossing on the way home after driving his daughter to school. This compared to around 1 minute on the way to school.\u201cI thought: Why in one direction it closes in a short period of time, and in the opposite direction it closes for a much longer period? So I decided to try and investigate what was happening.\u201dComing from an IT background, Guzman automatically thought there has got to be a technology solution to this problem. So he started building a simulated network and put his theory to the test.\u201cThe simulator measures the amount of traffic there is in the intersection, the average length of the queue, the longest queue, the number of stops within the queue, and so on.\u201cI spent many years on that issue just to see if I was wrong. I thought maybe I was, but I wasn\u2019t. You test the systems to make sure [they are] robust. Once all of that is done, then you start measuring the traffic flow.\u201cI started to actually simulate the current environment of the level crossing and then the proposed environment of the level crossing. Then I compared the results of both of them. And that\u2019s when I found I could reduce the traffic by about 50-70 per cent by moving the [level crossing] platform.\u201dGuzman\u2019s research was examined by professors of civil engineering at North Carolina State University and Queensland University, having recently obtained his PhD.His research paper \u2013 co-authored by his PhD supervisors Leslie Young and Konrad Peszynski \u2013 won a best paper award at the University of Melbourne and excellence award at RMIT University late last year.