I work on who gets exposed to what and where — using GPS trajectory data, network science and agent-based simulation.
📍 Pisa, Italy · ✉️ qrb.aliyev@gmail.com
- Visiting Researcher, MRC Epidemiology Unit, University of Cambridge (Nov 2025 – Dec 2026) — route choice under flood and heat scenarios using MATSim
- PhD in AI for Society, University of Pisa — defended 28 August 2026 with distinction. Thesis: Analysis and Routing-Based Mitigation of Vehicular Air Pollution Exposure
Vehicles choose the fastest route; pedestrians breathe whatever that route emits. My work estimates vehicular emissions from GPS traces, maps who is exposed near those roads, and asks whether routing can be optimised for health rather than only for speed.
- Emission estimation from GPS trajectories — a data-driven four-step process, validated against urban air quality measurements
- Spatial imputation — extending emission estimates to areas with no trajectory coverage
- Exposure-aware routing — joint optimisation of pedestrian and vehicle paths against a time/exposure trade-off, selecting among alternatives by hypervolume contribution
- MATSim route choice under flood and heat scenarios (Greater Manchester, synthetic data)
The honest version: most of my headline results come from constrained-route scenarios on synthetic city models, not measured populations, and the behavioural side assumes rational route choice without calibration. Each repo documents its own limitations rather than only the wins.
| Paper | Venue |
|---|---|
| Vehicle-Pedestrian Optimization Framework for Exposure-Aware Routing | Mobile Networks and Applications, 2025 (journal) |
| Optimization of Exposure-Aware Routing for Vehicles and Pedestrians | SSTD, 2025 |
| Exploiting Vehicular Data for Exposure-Aware Pedestrian Routing | IEEE MDM, 2025 |
| From GPS Traces to Individual Emission Exposure: A Data-Driven Four-Step Process | EAI INTSYS, 2024 |
| Analysis, Prediction and Mitigation of Exposure to Vehicular Air Pollution | SEBD, 2023 |
| Optimizing Exposure-Aware Routing for Vehicles and Pedestrians | CCS, 2025 (poster) |
| Repo | What it is |
|---|---|
| exawaro | Exposure-aware routing and behaviour simulation for pedestrians and vehicles on OpenStreetMap networks. Python (osmnx, networkx, shapely). Code from my thesis. |
| eide | Code for the four-step GPS-trace → emission-exposure process (INTSYS 2024). |
| matsim-project-gurban | Fork of upstream MATSim — my own experiment scripts for flood/heat route choice. |
Also here: MSc coursework notebooks (data mining, social network analysis, big data analytics) and a few forks.
Languages — Python (primary), R, C, Java, LaTeX; SQL (PostgreSQL, MySQL, MS SQL Server); Stata Modelling & simulation — agent-based simulation, routing optimisation, emissions estimation, network analysis, GIS/OSM, MATSim Data — spatiotemporal analysis, spatial imputation, trajectory preprocessing, ETL
Human languages — English, Azerbaijani (native), Russian, Turkish, Italian (intermediate), German (elementary)
MSc Data Science and Business Informatics, University of Pisa — 110/110 e lode. Erasmus semester, Vrije Universiteit Amsterdam (Computer Science, 2020–21). BSc Business Administration, ADA University, Baku — magna cum laude.
Grants: ISTI-CNR Grant for Young Mobility (2025, funded the Cambridge visit), PhD research grant in AI for Society (2022–25), merit-based admission grant (2018).