5. Concluding remarks
Estimation of travel time between a set of origins and a
set of destinations through a transportation network is a
common task in spatial analysis. Calibrating the O–D travel
time matrix in a commercial GIS package requires extensive
data collection and processing to prepare the road
network and also adequate knowledge of the software to
implement related tools. Both are no trivial efforts. This
research has developed a desktop tool to complete the
task by calling the Google Maps API. The Python program
automates the process by reading the layers of origins
and destinations in geographic coordinates, executing a
HTTP request to access the Google Maps and calibrate the travel time between each O–D pair, and saving the results
in an external ASCII file. By comparing the approach
with the commonly used ArcGIS Network Analyst module,
several advantages are identified: no need of preparing
a road network, using a more updated road network, and
accounting for congestion in high-traffic areas and peak
hours. Our case study in accessibility analysis indicates that
an accurate estimate of travel time is essential in spatial
analysis.