Smart cities, a digital twin, and how a city moves

python kafka deck.gl

imec was my longest engagement, and it moved through three quite different chapters.

It started with smart-city prototyping: research projects and the in-house tooling to enable them, with map visualizations of sensor data to help decision makers and citizens make informed calls.

Then I joined the Digital Twin team and moved toward the backend — building and maintaining a range of micro-services, processing event streams with Kafka, and picking up DevOps and infrastructure-as-code along the way.

The mobility model

The last chapter was a data-science project on urban mobility: understanding how pedestrians, cyclists and cars move through Flemish cities, and predicting what closing a given road does to neighbouring streets. I was involved end to end — data acquisition (static datasets and scraping web APIs), transformation and cleansing, modelling, validation and interpreting the output.

The model
A convex optimization program built on dynamics equations, solved with OSQP — turning messy movement data into something a city could actually reason about.
At a glance
clientimec — Leuven year2019 – 22 duration~3.7 years rolefullstack / data — freelance
Three chapters
smart-city tooling
sensor-data maps for decision makers
digital twin
micro-services, Kafka, IaC
urban mobility
convex optimization, end to end
Stack
typescript react redux redux-saga mapbox deck.gl graphql kafka docker kubernetes azure devops terraform python pandas osqp
want something similar?
Tell me what's slow, manual, or fragile
Start a project