Thread: I'm working on a new concept visualisation for occupancy data in BODs which utilises #3d 'mounds' to show the net change of people on buses. I thought I would share my janky workflow to give a bit more transparency into how this is made 1/
2/ First I start with 24 hours of data (Newcastle), whenever there is a change in occupancy levels points are created for the change. It's nice to see patterns along roads but it means very little at this stage, huge overlap in points etc.
3/ I filter points for each hour and individual change state (positive/negative). Hourly totals give nice patterns of commuter activity and are generally easy to manage. There is still overlap though and volume is difficult to make out at any given point.
4/ Each point is given a random offset along the z, immediately you can start to see the volume of people when vertical offsets are applied, less overlap, clearer message.
5/ Roughly speaking I want to create pyramidal distributions so when the spheres collapse they form mounds. So a rand float is generated for each point and its x/y position are offset and multiplied by it's inverse height value. Sort of creates a decent pyramid distribution.
6/ Spheres are generated at each point and a 'point relax' function spaces the spheres out so there is no overlap. Mounds begin to take shape but they need to be fall to a ground plane ideally. So I need gravity to take effect...
7/ Here is the fun part - using DOPs in Houdini allows you to apply rigid body dynamics to objects so they simulate real-world forces. Applying a load of friction to the ground plane and spheres means they settle without spilling too far, creating nice little 'occupancy piles'!
8/ This process for each hour split by +ve and -ve net change will form the basis of the next #animation to show the distribution and volume of where people get on/off the bus. These visuals are complex and take time and experimentation but the results are worth it! #dataviz
9/ And finally, ideas almost always come from inspiration and these occupancy piles were partly inspired by the hugely impactful "Drowning in Plastic" DataViz from 2019 (graphics.reuters.com/ENVIRONMENT-PL…)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
🧵 The Global Tanaka Project – Update 3 🌎🌍🌏
This week's update shows a busier globe with a soothing score 🎶, more plane/ship activity, biomes and vegetation, landmarks, volcanoes, a moon and the ISS satellite! Details below... #map#animation#3d
Biomes:
Using data from sciencebase.gov/catalog/item/5… I wanted to create a distinct feel for each ecosystem – this needs work still and I need to source some better low-poly vegetation models but I think it helps to distinguish eco regions.
Volcanoes:
I love @SonjaKuijpers idea about adding volcanoes and so found some data for volcanoes (volcano.si.edu via ArcGIS Online) with an ‘advisory’ or ‘watch’ threat level, otherwise there are too many! I need to get these animated and integrated properly…