Grab x Pulis
Detailed analysis done to investigate possible reasons for delay in Grab services for NUS Data Analytics Competition 2022, to be found in here and here.
Our main tech-stack:
- Vahalla, a C++ implementation for map matching.
- ipyleaflet, for very interactive visualizations of geospatial data analysis
- geopandas
- Dask
- matplotlib & seaborn
We've shortlisted the reasons to be:
-
Traffic bottlenecks at popular shopping malls due to narrow infrastructures of pickup points. We comparatively found out that pickup speeds at Changi Airport with optimized pick-up and drop-off pioints are much faster at the initial and end-timings of each trip, compared to popular shopping malls with narrow queues at their pick-up and drop-off locations.
-
Drivers picking inefficient routes, as we compare the actual driver routes taken with popular Google Maps and Open Street Map routes which we pulled using Google Maps API and
osmnx
. We found out that drivers's supposed "shortcuts" are more often slower, albeit, there were in-fact expert-curated routes which were actually even faster than Google Maps and Open Street Maps. These insights could be used to augment Grab-Nav!
Team:
- Keng Hwee Lead @kenghweeng
- Russell Saerang @RussellDash332
- Sean Gee Zhing @pikasean
- Terry Lim @terrylimxc
- Jonathan Chen @cysjonathan