Geographically-aware Transformer-based Traffic Forecasting for Urban Motorway Digital Twins
Krešimir Kušić, Vinny Cahill, Ivana Dusparic · 2026
Read Geographically-aware Transformer-based Traffic Forecasting for Urban Motorway Digital Twins by Krešimir Kušić, Vinny Cahill, Ivana Dusparic (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-05983.txt.
Description
A motorway digital twin needs more than live sensor feeds: to help managers act early, it must also predict traffic ahead. Krešimir Kušić and two co-authors build GATTF, a geographically aware transformer for that forecasting job. The model reads the geography between spread-out sensors through their mutual information, so nearby sensors that move together inform each other.
They test it on live feeds from Geneva's motorway network. Adding geographic awareness beats a standard transformer on accuracy, with no added model complexity. The gain matters for traffic control rooms, where a twin that sees ahead can warn of jams before they form. GATTF shows that where sensors sit is itself useful signal, not just background detail.
Catalog record
| CREATOR | Krešimir Kušić, Vinny Cahill, Ivana Dusparic |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-05983