Dashboard Week Day 3 - Bike collisions in Seattle

Day 3 saw us building dashboards based on Bicycle traffic data. Full details here: https://www.thedataschool.co.uk/andy-kriebel/ds28-day-3

My plan was to add on collision data, and join in via a spatial match in Alteryx - ultimately I couldn't get this to work in the time we had but with a bit more tinkering I think I could. In the end I used a Tableau relationship model.

Here was my initial plan:

The plan included time series charts showing cycling trends over time, a map view with some BANs and some bar charts. Ultimately my final design followed this quite closely.

Here is my Alteryx flow:

The main part involved unioning all the traffic data and removing pedestrian traffic. I also aggregated the data up to a day level instead of an hourly level as the collision data was a bit of mess with the times so kept it simple for this analysis.

Here is my final dashboard:

https://public.tableau.com/views/SeattleCycleTrafficCollisions/CyclingTrafficandCollisions?:language=en-US&:display_count=n&:origin=viz_share_link

Ultimately is follows most of my plan, although there isn't any date filter as the relationship model wasn't allowing me to filter by date but I think with a little tinkering I could get this to work.

Author:
Jacob Kilroy
Powered by The Information Lab
1st Floor, 25 Watling Street, London, EC4M 9BR
Subscribe
to our Newsletter
Get the lastest news about The Data School and application tips
Subscribe now
© 2025 The Information Lab