The team explored the potential for increased use of low carbon transport modes in a suburban context. The outputs highlighted the interdependencies between modes, and the factors that are likely to encourage greater use of active transport and public transit.

What the model describes

The linked model presents a ‘toy’ version of the ABM for Transport Mode Choice (TMC). This model is used to describe uptake of low carbon transport modes in a suburban precinct under different scenarios in a way that considers realistic models for commuter behaviour, and their responses to supply interventions that improved accessibility by active and public transport.
Please note that the Netlogo Web model was inserted as it was exported from the desktop application. Improvements to the user interface will be explored in a future phase of this project, which will include spatial visualisation.
Link to model

How we are modelling it

The ABM-TMC was implemented using a simple agent-based model (ABM) within the Netlogo (Version 5.3.1) environment, and applied a Consumat approach to simulating travellers’ mode choice. The Consumat theory was used to simulate for each time-step travellers’ decision-making based on satisfaction of travel needs in the previous step, and the level of uncertainty. The following decision modes were simulated: repetition (satisfied and certain), imitation (satisfied and uncertain), optimisation (unsatisfied and certain) and inquiry (unsatisfied and uncertain). The decision making profile was modelled for the population travelling into the case study location based on analysis of survey data (described below). This was then validated against actual travel behaviour using ABS census (2016) data.
As an input into the Consumat model, we developed multinomial regression approach based on survey data to describe how intention for transport mode choice is influenced by behavioural drivers. The variables from the survey data that were found to be statistically significant in explaining the use of active travel modes and public transit were:
  • Travel time
  • Number of public transit connections between home and work
  • Public transit convenience (frequency of service and distance to stop)
  • End of ride facilities availability
  • Car practicality (based on access to car and parking availability at work)
  • Intrinsic priorities (combined metric indicating importance of comfort, health, environment and flexibility)
In addition to the survey, other input data used to characteristic the case study included:
  • Projected population
  • Employment projections
  • ABS census journey to work data
  • Land use map
  • Transport networks

Running the model

To run the model, the user selects a commuter file, which can segment the travelling population to be modelled into the following groups:
  • General population (all)
  • Younger people (<25 years)
  • Older people (>65 years)
  • Women
  • Men
The user then selects to intervention scenarios, which are:
  • Upgrading of off-road active transport paths connecting origins and destinations (including transit nodes).
  • Improved public transit services that reduced number of connections and frequency of services between origin and destination points.
  • Combination of both upgrading of off-road paths and improved transit services.
The model runs over a 10 year period (to 2028), with the following outputs:
  • Change in modal split (percentage)
  • Change in car dependence (percentage reduction in people travelling by car)
  • Average per capita emissions