Strava Metro : City of Bath Case Study

CycleBath in partnership with Bath Hacked approached Strava for the City of Bath Metro Data set. This documents the journey we have been on.

Transcript

1. Strava Metro City Of Bath Adam Reynolds @awjre awjreynolds@gmail.com IT Consultant by day Spend my evenings/weekends as: CycleBath Chair Cycling UK BaNES Local Campaigner Member of Bath Hacked
2. CycleBath
3. Bath Hacked • Joint council/community Community Interest Company • Putting open data and smart thinking at the heart of the city. • Bringing bright people and quality data together to do useful things for the community.
4. What is Strava? A website and mobile app used to track athletic activity via GPS
5. http://labs.strava.com/
6. What is Strava Metro? Enables deep analysis of cyclist and pedestrian activity.
7. Timeline • January 2015 CycleBath in partnership with Bath Hacked approach Strava to get Bath Metro data set for hackathons • February 2016 Contracts signed & data delivered • March 2016 • Bath Hacked & CycleBath publish initial analysis • CycleBath engages local cycling community to understand data better • May 2016 Bath Hacked releases Strava tool
8. Such Wow Much Data 129,568 activities in 2015 12,139 Strava users 80% of Strava users in Bath are male 50% of activities are commutes 1.3GB of data
9. Athlete Demographics
10. Athlete Activity - Days
11. Athlete Activity - Hour
12. 2015 by month and hour
13. City Topography
14. Cyclists and cycle paths
15. Visualising key routes
16. Cycled routes and intersections
17. Key commuter areas
18. Animating the data
19. Seeing patterns
20. It’s all very nice but… • Image and video generation is time consuming • Difficult to play with the data • Developer Cost
21. strava.bathhacked.org
22. Layers • Athletes Total • Athletes Average • Forward Reverse Athletes Total • Commute Total • Forward/Reverse Speed • Junction Totals • Junction Maximum Wait • Junction Median Wait • DfT Cycling Collisions
23. Layers have complexity Quantile Categorisation
24. We can zoom in
25. We can look at behaviour
26. Analysing Junctions
27. DfT Cycle Accident Data 1996-2013
28. Can we trust the data? A 12 hour towpath manual pedestrian and cycle count on Wednesday 1st July 2015 showed usage was pedestrians 682 (48%) and cyclists 737 (52%). Strava reports only 4100 for the year!
29. Real World Application
30. Strategic Network Review
31. Sydney Gardens Heritage Lottery Bid Workshop
32. Use within the council • Early Days • Exploring use as part of path maintenance regimes • Will be used for helping develop future cycle infra • Already feeding into Public Realm consultations • Enables better prioritisation of resources
33. Other Tools? http://pct.bike vs strava
34. What we didn’t do Origin -> Destination Analysis
35. Considerations for use • Cost is $0.80 per rider in a year. • Bath @ 12,139 riders ~ £10k 29 km² 89,000 population • Bath and North East Somerset cost ~ £20k 351 km² 180,000 population • LA boundaries are not good selectors of data. • It is only one truth. Local knowledge essential. • Can give false positives dependent on road works in that year. • Local IT developer skills (ArcGIS) • Public tools cannot be too time granular due to Strava Contract.
36. Any questions? If you want an interactive Strava Metro Data tool we can help point you in the right direction or you can commission us to do it for you. hello@bathhacked.org www.bathhacked.org