Toronto Evacuation

This is a simulation of the evacuation of the City of Toronto in case of an emergency. An optimal spatio-temporal evacuation (OSTE) model is used to optimize the evacuation time and it was concluded to be 4 times faster compared to do-nothing strategy. The simulation shown in the video has considered different modes transportation in the evacuation which includes the vehicular traffic, and rapid transit and mass-transit shuttle buses.

Toronto Evacuation - 2

This video is a simulation of the evacuation traffic flow coming from the City of Toronto into the Grand Toronto Area in an emergency situation. The simulation is created using DynusT mesoscopic Dynamic Travel Assignment model.



This video demonstrates the application of Multiagent Reinforcement Learning for Integrated Network of Adaptive Traffic Signal Controller (MARLIN-ATSC) on one of the intersection traffic control system. Each traffic signal controller learns from other controllers with objectives to alleviate the traffic congestion. A simulation has suggested that by employing MARLIN-ATSC average intersection delay of the traffic network can be reduced by 39% and travel-time can be saved by 26%.


This is a video showing the simulated traffic flow in the Toronto waterfront area, which includes all roadways south of Queen Street and in between Humber River in the west and Port Area in the east. The dynamic simulation is developed with the Paramics simulation software. As part of the QueensQuay Traffic Feasibility Study, micro-simulations for the traffic in the waterfront area are made to facilitate finding solutions to the waterfront access challenge that the elevated Gardiner Expressway brings.


This video is a simulation of the traffic flow of a section of Highway 401 going eastbound during the morning peak period of July 6, 2011. The simulation compares the original traffic flow (without ramp control) to the simulated traffic flow that uses reinforcement learning based ramp metering (RLRM) as the control mechanism for the merging traffic from the ramp. The total travel time for transportation users were found to be reduced by 40% with RLRM.

St Clair


This video is a simulation of traffic on a segment of highway 401. It’s made to better visualize the status quo and explore the potential impact of adding exclusive trucklanes on the highway.


This video shows how traffic at an intersection can be improved by introducing Traffic Signal Priority (TSP) to the traffic controllers. The real time signal control is optimized using genetic algorithms.