Written by Erin Morrow (Senior Consultant at Arup) & Peter Debney (Senior Consultant at Oasys Software Ltd)
Union Station is Toronto’s busiest transportation hub. Situated in the downtown district, it is the main terminus for the GO regional rail network, a connection into the city’s streetcar and subway networks, plus a gateway into the PATH system of walkways.
The station currently handles over 30,000 passengers during the morning peak hour and more than 250,000 passengers over a typical business day. These numbers are expected to more than double over the next ten years, including over 70,000 passengers during the peak morning hour in 2021, as GO Transit’s expansion program is put into operation.
Assessing pedestrian demand
Due to the predicted increase in passenger numbers at what is already a very busy location, the historic station building requires refurbishing. Its passenger handling capacity also needs expanding, and the surrounding streets must be improved to make them safer and better for both pedestrians and cyclists. The City of Toronto’s master plan was adopted in 2004 and Arup was awarded the contract to assess the pedestrian flows through Union Station during refurbishment and in the predicted conditions of 2021, thus informing and supporting the planning and design of the revised station.
The first phase was to survey the peak pedestrian conditions in and around the Union Station and use these to create quantitative forecasts for passenger flows both during renovation and in 2021. This analysis focused on the peak hour and peak 15 minute pedestrian volumes during the morning and afternoon peaks, plus special events at the Air Canada and Roger’s Centres. The study revealed a number of opportunities and constraints for the refurbishment project. It also gathered data for the second phase, which was for detailed pedestrian simulation of the station to check pedestrian flows and discover the best locations of retail, commercial and transit-related facilities.
Arup’s work was required to answer four fundamental questions:
1. How would the proposed Concept Plan support or otherwise impact pedestrian flows?
2. Was the Union Station Concept Plan appropriate from a pedestrian flow perspective?
3. What are the internal and external congestion points, and what conditions may be causing congestion?
4. Where are areas of flexibility that provide opportunities for other precinct and station revitalisation initiatives?
This planning study involved the creation of agent-based simulation models of the station and surrounding streets. To ensure that the model results were accurate, the current conditions in the station were modelled using the survey information from the first phase, including minute by minute breakdown of where the pedestrians enter the study area (origin) and to what destinations they were going to. Each agent was assigned a train platform either as an origin or destination, with distribution based on the train schedule. The resultant pedestrian model was then calibrated against actual flows within the station and adjusted so that the analysis was consistent with observations.
This model was then modified to analyse the 2021 predicted passenger flows and to also check each stage of the refurbishment to ensure that the station would continue to function while the rebuilding work closed off parts of the concourses.
Having previously tried the existing pedestrian modelling packages on the market and found that they were insufficient for its purposes, Arup created its own program, MassMotion, to address the shortcomings of the available programs. MassMotion is based on intelligent agents, a full 3D model of the environment, and John Fruin’s industry standard planning and design guidelines for pedestrian behaviour.
The principle differentiator between MassMotion and other pedestrian simulation tools is that it actually models pedestrian behaviour rather than testing a user’s preconceptions about pedestrian behaviour: it is non-deterministic and emergent. This means that the individual agents in the simulation make their own choices about appropriate actions based on the dynamics of their environment and their action’s effect on other agents.
For example, if a room has doors on all four sides, other pedestrian simulation tools require the user to input what percentage of the room’s population will use each door. In a MassMotion simulation each agent decides which door to use based on what it knows about the distance to its goal and how long the queue is for each door.
There are two noteworthy advantages to such a system:
1. A MassMotion agent only needs be assigned an origin and a destination: it then navigates its own way through the environment without requiring the user to input percentage splits at all potential branching points and for all sub groups within a simulated environment. This means that the more complex an environment becomes, the quicker it is to set up or modify a MassMotion model in comparison to other programs.
2. Unlike other tools, MassMotion actually predicts how rational pedestrians will navigate through an environment and how they will respond dynamically to changing conditions within the simulation: for example agents will change their route if congestion becomes too great.
The 3D MassMotion environments are created using CAD tools such as AutoCAD or SketchUp and are based on architectural drawings or imported direct from BIM models. The information contained in the drawings is used to create a number of polygons representing the walkable areas of the environment, plus solids representing obstacles within those environments.
The walkable areas are further broken down into circulation spaces such as floors, pavements, and platforms, and into connection elements such as doors, stairs, escalators, etc.
The arrangement of circulation areas and connection elements forms an implicit network, defined by geometric proximity, which defines possible route permutations for the simulation agents to navigate. For the Union Station environment, the platforms and station were modelled in detail, while the network of adjacent streets and walkways were modelled within the same file in a more abstract fashion.
After extensive simulation of the proposed concept plan, Arup concluded that its proposed configuration would support the estimated pedestrian volumes for 2021. Also, while there will be some areas that experience high densities of commuters during the morning peak 15 minutes, the traffic should continue to flow. Likewise the external pedestrian routes were providing sufficient capacity with just a few exceptions. The model also revealed some underutilised areas within the station. By adjusting stair, retail and service locations, the team developed layout modifications to improve the balance of flows as well as the user experience.
Tim Laspa of Toronto City planning department spoke highly of the project.
“The pedestrian planning initiatives at Union Station have broken new ground for the way the city of Toronto will plan for pedestrian activities in the future. From a modelling and simulation perspective, the Phase 2 study was an exceptional test of the MassMotion toolset that Arup has developed. The very high volumes of pedestrians being simulated in combination with the complex layout of the station facility demonstrated the usefulness of an agent-based approach to pedestrian simulation and analysis.”
Arup has continued to develop MassMotion by improving the software based on lessons learned from simulation work done for the City of Toronto at Union Station, with improved model creation tools, BIM interfaces, and additional functions such as evacuation events for fire engineering. MassMotion is now commercially available.