School walking-route impact, in hours instead of days
Walking times for a whole school catchment, recalculated live as you change the roads.
The problem
Traffic teams assessing a school need to know how long students across a whole catchment take to walk in, and how that changes when a road closes, a new road opens, or the school moves an exit. It sounds simple until you try to do it at scale. The usual way is one address at a time in an online map: type it in, drag the route to model the change, note the result, move to the next. For a whole catchment that is hundreds of lookups, and it runs to three or four days. Test a second scenario and you start again.
What I built
The team works across a region where the government publishes the spatial data, so I built them a map downloader: they browse to wherever they are working, show the lots and the roads, and pull the data down, accurate from the source.
From there, a workflow classifies every home around the school by its shortest walking route. The script behind it is large, but the input is a single button.
What makes this work
Delete a road, add a school exit, move the school itself, and the whole catchment recalculates in front of the team as they watch. Testing a scenario stops being a repeat of the entire exercise and becomes a single change, because the model stays live rather than a one-off calculation.
What changed
What took three or four days now takes a couple of hours, and it comes with something the manual way never had: a live model the team can push on. Block a road, try a new exit, watch the catchment redraw, instead of starting the whole assessment over.
Notes on maintainability
It started deliberately lightweight, inside a tool the team already used, as a cheap way to prove the workflow before it grew into something more.