Hi! I am a Senior Researcher in the Ecological Systems Design group at ETH Zurich, working on transport, land use and sustainability, with one question in mind: how does where people can go shape how they travel, and what does that travel cost and give us?

My work follows that question along a chain. I measure the built environment and accessibility at the resolution at which people experience it, for every mode and across whole countries. I use that measurement to understand travel behaviour, from aggregate demand to the choices people make between modes and destinations. And I carry the results through to outcomes: emissions and energy use, but also cost, access, and who gains and who loses, because a mobility system is there to give people reach, not only to be clean. I came to transport from the impact side, trained in environmental engineering and in data science and policy at MIT’s Institute for Data, Systems, and Society; the methods are those of transport modelling, spatial analysis and econometrics, the purpose is to know what the travel does.

Models should be as complex as necessary, but as simple as possible, and open enough for others to check and reuse. Much of what I build is therefore public: aperta, a Python library for calibrated, multimodal accessibility analysis, the national datasets [1, 2] built with it, and tools such as carboncounter.com. You’ll find them under Tools and Data.

In my spare time, you'll sometimes find me behind the viewfinder of a camera, especially when travelling. I've allowed myself to share a couple of pictures. Enjoy!

Marco Miotti

Researcher (Oberassistent), Lecturer | Transport Planning · Urban Analytics · Impact Assessment

Chair of Ecological Systems Design
ETH Zürich

marco@miotti.me CV

Tools and Data

aperta

aperta is a Python library for multimodal accessibility analysis at any spatial resolution. It routes walking, cycling and car trips on OpenStreetMap networks and calibrates them against travel survey or GPS data, so that the typical experienced speed, junction delay and congestion along every edge are fitted to observed door-to-door travel times rather than assumed. On top of the calibrated networks it computes accessibility as cumulative counts, nearest-k costs and gravity measures, in distance, travel time or generalized cost; the generalized costs come from a mode-choice model, which makes accessibility comparable across modes and lets it be combined into a logsum. It is written for researchers and planners who need accessibility measures they can reproduce, recalibrate and transfer to a new region.

aperta is currently prerelease; expect a first mature version with a stable API at the end of 2026. The pipeline built on it, aperta-atlas, is described below.

National accessibility surfaces and underlying calibrated networks

I've built an end-to-end accessibility pipeline based on aperta called aperta-atlas. For Switzerland, it turns public inputs (OpenStreetMap, the federal population and employment registers, and the national transport model) into two open datasets: calibrated walking, cycling and car networks, whose speeds, junction delays and congestion are fitted to observed travel times from the national travel survey and GPS tracking, and accessibility surfaces for the whole country on a hexagonal grid of about 15,000 m² per cell, for 18 destination types and seven mode profiles including e-bikes and reduced-mobility walking. Accessibility is given in distance, door-to-door travel time and generalized cost, so that modes can be compared and combined into a logsum.

All fitted coefficients are released with the data, so the surfaces can be reproduced from public inputs alone, recomputed for specific population groups, or run for another region where OpenStreetMap exists. Both datasets are prerelease (v0.2.0); a data descriptor is in preparation, and a first stable release will follow with aperta's.

Carboncounter.com

Carboncounter.com is a website that lets users compare the lifecycle emissions and costs of new cars against other cars and against climate targets. In 2023, I updated it with more recent electric vehicle model information.

A European version, developed in collaboration with myenergy Luxembourg, can be found at carboncounter.lu.

I built carboncounter.com using AngularJS and D3.js as an interactive version of figure 1 in our Environmental Science & Technology article.

In the News

Publications

Peer-Reviewed Journal and Conference Publications

Book Contributions

Technical Reports

Photography

© 2016-2026 Marco Miotti. marco@miotti.me