Mobility On-Demand Laboratory Environment

The provision of Mobility as a Service (MaaS) is challenging the balance between private cars and public transport. Ride-sharing services have demonstrated car ownership can decline when travellers’ needs are satisfied. Moreover, pop-up mobility services are flourishing around UK cities, providing a useful asset for Local Authorities (LAs) and extending the provision of public transport services.

The Mobility on Demand Laboratory Environment (MODLE) project aims to demonstrate how mobility service providers can run profitable services creating opportunities for LAs and Public Transport operators in providing better service for users. To achieve this goal, the Buzz mobility service, operating in Bristol (UK) will be fully supported by an Agent Based Model (ABM) developed to inform the service operations and to increase the performance of the service.

MODLE is experimentally developing affordable, direct, spontaneously available, commercially sustainable, door-to-door transport services (delivered in shared vehicles) which compete with the convenience and cost of the car, thereby reducing congestion and improving access to employment and services for those without one.

The Buzz service is an e-hailing minibus service that will improve mobility in an area with poor accessibility and strongly relying on private cars. The catchment area of the service is north of the river Avon and north of the A420 in the east, however, the model extends to Greater Bristol, with comprises Bristol City, South Gloucestershire, North Somerset, Bath and North East Somerset. Focus is given to new and regeneration areas (Filton, Avonmouth and Severnside), where level of congestion and lack of parking are affecting the quality of life of users.

In order to assess the introduction of demand response services a MATSim simulation was employed. In the last decade, the shift from using typically aggregated data to more detailed, individual based, complex data (e.g. GPS tracking) and the continuously growing computer performance on fixed price level, leads to the possibility of using microscope models for large scale planning regions.

MATSim is a multi-agent micro-simulation model. In MATSim each modelled agent (person, vehicle, etc) contains its individual settings. The sum of all physical agents should reflect the statistically representative demographics of the region.

In the context of a rapidly expanding population and technological advancements, a new range of possibilities are opening up to more flexible, demand-responsive, safe and energy efficient transport services. To date, no existing solutions provide large-scale microscopic simulations that include all the above components.

MATSim is addressing the issue of demand-responsive design by introducing the dynamic vehicle routing and scheduling contribution. The current development can model supply and demand to optimise fleet operations in the context of a large scale microscopic simulation.

The dynamic vehicles extension is responsible for listening to simulation events, monitoring the movement of the vehicles finding the least-cost paths, computing schedules for drivers/vehicles and coordinating cooperation between driver, passengers and dispatchers.

MATSim supports very detailed public transport modelling; transit vehicles run along the defined transit line routes, picking up and dropping off passengers at stop locations, while monitoring transit capacities and maximum speeds. Data used to simulate public transport in MATSim can be split in three parts:

  • Stop locations: 45,995 entries
  • Schedule, defining lines, routes and departures
  • Vehicles: 95,454 buses
  • The model contains 700,000 private car journeys.
  • Adding mobile phone data in the simulation improved the realism of the model, new travel patterns have emerged.

Results

The MODLE project accessed the impacts of the proposed ride-sharing service, specifically through a modelling exercise:

  • Reduction of congestion.
  • Improvement in accessibility to employment and services.
  • Increased number of longer journeys made sustainably.
  • Improved accuracy of local authority interventions.
  • Transferability of the PT Service to other cities.

Next steps

Introducing the dynamic agent’s extension in the simulation in order to build a demand responsive model which can supply fleet operations in the context of a large-scale simulation.

Deploying the e-hailing service for the study areas.

Cookies on Catapult explained

To comply with EU directives we now provide detailed information about the cookies we use. To find out more about cookies on this site, what they do and how to remove them, see our information about cookies. Click OK to continue using this site.

OK