Background¶
The problem¶
Roughly 1.1 billion people live in informal settlements — neighbourhoods that grew without a planned street layout. When homes are not reached by a road, they are also not reached by the services that ride along roads: ambulances and fire trucks cannot get in, and piped water, sewer drainage, and power have no corridor to run through. Reblocking is the practice of adding the least-disruptive set of new roads that reconnects every parcel to the existing street network, displacing as few homes as possible.
Why roads matter¶
| Function | What a road brings to a settlement |
|---|---|
| Emergency access | Faster ambulance, fire, and disaster response reaching every home |
| Utility corridors | Space to run water, sewer drainage, and power infrastructure |
| Accessibility | Everyday mobility for residents — access to schools, jobs, and services |
A reblocking proposal is only useful if it buys these gains cheaply. That is the tradeoff this project measures: the access a road network unlocks, against the homes it displaces to do so.
Prior work¶
This project builds directly on two lines of research from the Santa Fe Institute and the Mansueto Institute.
Optimal reblocking. Brelsford, Martin & Bettencourt (2019) framed reblocking as a formal optimisation over a block's parcel graph. Their method is optimal but its worst-case running time is exponential in the number of parcels, so it cannot run on large blocks, and it is inherently single-block: each block is solved in isolation. The methods here run efficiently and scale to regions made of many large blocks, which lets added roads stay continuous across block boundaries rather than stopping at the edge of each separately-solved block.
Detecting where to reblock. Soman et al. (2020) detect informal settlements worldwide by a
topological analysis of crowdsourced maps — but their per-block Voronoi tessellation is
prohibitively expensive to run at metropolitan scale. This project instead screens with
depth_density_proxy (√(nA)/P · n/A), a cheap heuristic that rewards blocks which are both deep
and crowded — built from nothing but each block's building count, area, and perimeter, with no
Voronoi and no peel — scoring every block in an entire metro in a single fast sweep. What it finds is
checked against the City of Cape Town's own informal-structure survey, and reported on the
screen bake-off page.
References¶
- Brelsford, C., Martin, T., & Bettencourt, L. M. (2019). Optimal reblocking as a practical tool for neighborhood development. Environment and Planning B: Urban Analytics and City Science, 46(2), 303–321.
- Soman, S., Beukes, A., Nederhood, C., Marchio, N., & Bettencourt, L. M. A. (2020). Worldwide Detection of Informal Settlements via Topological Analysis of Crowdsourced Digital Maps. ISPRS International Journal of Geo-Information, 9(11), 685.