Geospatial simulator for urban transportation

Client

Blaise Transit

Funding

Partnership Commitment Grants (PCG)

Objective

The project aimed to help Blaise Transit improve and automate its platform using a geospatial simulator for the selection of on-demand urban transit corridors, and to perform a cost-benefit analysis of service provision. The main objective is to investigate and propose an approach to aggregating and analyzing ridership data to determine when and where on-demand services will present optimal results.

Methodology

The project introduced a series of measures to help identify a pilot area for the operation of an on-demand public transport network. Various methods were implemented using Python.

Next, we prepared bus demand datasets over several days and enriched them by integrating them with the ridership dataset and distance matrix data.

Using the enriched demand data, we group transport requests by adapting k-means clustering, separating requests within each group and preparing input data for the dynamic routing process.

Finally, we simulate the result by applying dynamic routing to each group’s requests. We measure the performance of dynamic routing by applying a set of metrics to the routing results and comparing the results to static traffic.

Software

Python
GDAL
QGIS

Domain

Transport and logistics