Context
Cities produce enormous amounts of data and expose almost none of it usefully. Infrastructure, mobility, and public services generate signals that mostly evaporate.
Problem
Public data is fragmented, inconsistent, and rarely connected to a model of the place it describes. Without a coherent twin of the system, the data stays trivia instead of insight.
What I did
- Experimented with mapping infrastructure and public data into a coherent, queryable model of a place.
- Treated the city like a machine worth instrumenting, the same diagnostic instinct, scaled up.
What I learned
- The 'make the system explain itself' problem is not unique to computers. It shows up anywhere telemetry outruns interpretation.
A city is a machine, just a very large, very messy one. This set of experiments applied the same instinct I bring to computers. Instrument the system, model it, and turn raw signals into something a human can reason about.
Why it is here
It is a side thread, but a revealing one. The pattern that drives my main work, telemetry is abundant and interpretation is scarce, turns out to be nearly universal.