Navigation path

Decrease textIncrease textDividerPrint versionRSSDivider

Advanced Virtual Agents for Testing the Accessibility of Rail Stations

AVATARS will develop a range of innovative technologies to simulate passenger movements. The tools will enable a population of structures with large groups of realistic, interacting agents in a series of realistic scenarios. To achieve this, AVATARS will develop three major software components. The evaluation of these tools will be carried out under a series of case studies, developed throughout the project.

Tags: Rail


Current rail transport infrastructure design is led by aesthetics and prescriptive regulations. Better terminal design would increase passenger throughput and reduce overcrowding but it is very difficult to design for performance without accurate simulation tools. Existing tools are available but none specifically for general circulation movements.

AVATARS falls under the Objective “Advanced Design and Production Techniques”. Within the overall objectives of area 2, the project addresses the need to “reduce manufacturing costs by 30-40% and production lead times by 25%” by the development of a tool to provide ‘right first time’ manufacturing of multimodal transport infrastructure. Specifically, AVATARS addresses research domain 2.7 by providing “design technologies to improve vehicle interfaces with transport infrastructure and other vehicles/vessels from the same and different transport modes”. This is accomplished through the development of simulation tools to optimise the comfort and flow of passengers within the transport interchange environment. These tools will be developed to consider not just the terminal but also the area immediately outside the terminal, covering interchanges to other modes of transport. Linking with the overall aims of the objective, AVATARS will assist in developing new construction concepts for transport infrastructure, specifically improvements in safety, comfort and efficiency. Lastly, the project will examine issues to provide design techniques used in vehicle infrastructure creation aimed at developing safe and comfortable products with reduced operational cost.


The principal objective of AVATARS is to develop a simulation tool to provide designers with a useful performance measure for train station designs which in turn will be able to improve their design based on this feedback. The improvement of a terminal design can be measured in four ways:

  1. Improvement in the station’s ability to function in unusual or emergency situations
  2. Improvement in flow rates around the station
  3. The ability of the station management to organise the movement of passengers so as to provide easy means of access while exposing them to the facilities available
  4. Reduction in the number of negative emotions experienced by passengers.
These measures will allow for a further level of functionality to be built into an otherwise well-designed rail terminal. In order to achieve the above objectives, a number of quantifiable sub-objectives must be completed:

  • AVATARS will develop a scenario generator capable of creating a population of 500 unique agents.
  • AVATARS Agents will be capable of interacting in realistic groups of up to 20 individuals.
  • The agents will be capable of handling up to five consecutive objectives in one session, including both pre-set objectives and dynamic objectives emerging from the simulation.

Description of work

The AVATARS tool will innovate in a number of distinct areas:

  • Circulation-specific simulation tools covering new approaches to information dissemination and group communication
  • Unique agent behaviours linked to internal emotional states
  • Innovative scenario generation facilities for rapid prototype evaluation
  • Intelligent software to assist users in identifying the root causes of terminal performance issues.

The AVATARS consortium will be developing innovative technologies and, consequently, intends to protect its developments through the use of patents. The desire to patent work means that some of the deliverables for the project will be disseminated sparingly until the information contained in them is protected.


Below is a list of the deliverables that will be produced by the end of the project:

  • Report on data requirements and collection design
  • Report on agent interaction requirements and specification
  • Report on support tools requirements and specification
  • Report on case studies data collection
  • Report on exodus case analysis
  • Report on case studies definition
  • Agent interaction prototypes
  • Scenario generator prototype
  • Advisory module prototype
  • Report on testing procedures and quality assessment criteria
  • Report on test results and stage 2 recommendations
  • Final agent interaction prototypes
  • Final scenario generator prototype
  • Final advisory module prototype
  • Functionality report and further development plan
  • Final integrated prototype
  • Final evaluation report