Python Machine Learning Conference & GeoPython 2020

»Python, Let's Go Home. Quickly.«
2020-09-22, 11:30–11:50, Room 1

When your daily commute, shopping tour, or visit to family or friends offers many possible routes and real-time online navigation services are of no use for your individual mean of transport, you can still learn to find the optimal route with Python.

Online navigation services are usually bad at finding the optimal route for frequent journeys for people moving around the city. Every pedestrian and cyclist behaves individually and their time needed to get from A to B depends on their speed, acceleration, braking, collision avoidance, and general way of solving traffic situation.

I wanted to find the best commuting route, so I started tracking all my bicycle trips and wrote pygohome that analyses my GPX tracks and builds a graph of road segments and calculates statistical estimations of time needed to get from A to B between all my points of interest in the city and surroundings.

This is a completely personal data analysis that works anywhere for anyone, but only for the own tracks of the respective user.

In my talk I'll speak about: - History of the city and transportation - Navigation for different means of transport - Problems of measuring and estimating of the speed of individual transport in a city - GPX logging and analysis - Network model of city routes - Different types of intersections and their throughput for individual transport - Evaluation of results - Possible further steps

I'd love to discuss this topic with the audience and get more ideas/feedback from potential users.