LEGION Simulator

Unbounded Movement Choice

• Model space is continuous, rather than structured, based on a pixel grid
• Select any vector when optimizing step choice to satisfy individual preferences and objectives in the context of changing physical constraints
• This approach follows a two-stage process:

  • Macro-navigation

    • The selection of an entity’s desired direction to its next target, from its current position
    • Represented by a sequence of intermediate, focal targets that trace the shortest path from an entity’s location to the place where it leaves the model (or reaches its final destination)

  • Micro-navigation

    • Uses advanced, proprietary, artificial intelligence algorithms to apply micro-navigation to its entities, within a simulation
    • Algorithms enable entities to exhibit realistic pedestrian movement
    • Entities have an area of perception that adjusts dynamically, based on instantaneous information and accumulated memories
    • Entities assess information to decide their best immediate step
    • Micro-navigation algorithm takes several important considerations into account, including:
      • Early detection and avoidance of physical obstacles
      • Accommodation of personal space, preferred speed and more
      • Maneuvering to avoid collisions
      • Learning from accumulated memories
      • The ability to adjust individual preferences and attributes

Intelligent Entities

• Social, physical, and behavioral characteristics are assigned probabilistically from established profiles
• Social characteristics include gender, age, culture and pedestrian type, all of which shape typical movement preferences
• Physical characteristics determine body sizes
• Behavioral characteristics include memory, adaptability, and preferences for unimpeded walking speeds, personal space, and acceleration

Output and Analyses

• Numerical and graphical outputs derived from a LEGION Simulator model and user-defined areas, interrogated to provide user-defined combinations of these, based on user-defined thresholds
• Measure and quantify the performance of a site with respect to experience, circulation, and safety
• Key metrics include counts, flows, distances, densities, journey times, speeds, levels of service, and more
• Heat maps provide intuitive overviews to help identify areas requiring deeper analyses, including line graphs, histograms, stacked histograms, cumulative data, or even raw data, which can then be used for statistical analyses

LEGION Model Builder

LEGION Model Builder allows you to create predictive models of how a space will be used.

• Import architectural drawings (CAD) to define physical spaces available for pedestrian use
• Specify the anticipated pedestrian demand for the space
• Designate areas where interim activities such as queuing or waiting occur
• Link operational data to the model
• Plan routes and automatically lay out navigation maps
• Export model files for simulation and analysis in LEGION Simulator

LEGION Simulator API

• LEGION Simulator API allows users to add pedestrian models to external applications. One existing application is the integration with traffic simulation software Aimsun Next.
• Two interaction methods are possible:
  • Aimsun Next provides pedestrian infrastructure elements modeled using the LEGION Simulator API. This minimizes user effort to include pedestrians in Aimsun Next models. LEGION modeling experience is not required. Models are built separately, but the background geometry must be the same; crossings and other interaction areas must be built in both.
  • For users experienced with LEGION modeling, any complexity of LEGION model is possible.