What is it about?

We present a way to screen the earthquake vulnerability of unreinforced masonry (URM) building façades using only ordinary exterior photos, so assessments can scale beyond slow, expensive, on-site engineering surveys. URM buildings, made from brick, stone, or concrete blocks, are widespread and are associated with a significant share of earthquake fatalities, especially in places with limited monitoring and resources. The idea is to turn images into a physics-based estimate of how a façade might crack or weaken during shaking. Starting from multiple photos taken around a building, the method reconstructs a simplified 3D model and uses computer vision to identify visible features such as the façade shape and openings such as doors and windows. Those visible elements are then converted automatically into a structural model made of larger “panel-like” pieces (macroelements) connected in a way that can reproduce realistic damage patterns. The model is pushed through nonlinear earthquake-style loading in simulation, producing an interpretable output that highlights where damage is most likely to concentrate and how the façade deforms under cyclic loading. Importantly, the results are predictive rather than diagnostic: they are meant to help prioritize where expert inspection and retrofit attention should go, not to replace a full engineering evaluation. Because the approach relies on external imagery, the paper also points toward future city-scale use with widely available street-level images, potentially enabling broader, faster risk mapping across neighborhoods and borders.

Featured Image

Why is it important?

Low-rise unreinforced masonry buildings are still widespread and are linked to a large share of earthquake fatalities, especially in places that lack dense early-warning and monitoring infrastructure. As cities grow, that creates pressure for proactive, scalable ways to identify which structures deserve urgent attention, before the next earthquake turns uncertainty into loss. The work tries to close a longstanding gap between two worlds that rarely meet at scale: fast scorecard screening and high-fidelity structural simulation. Existing rapid visual screening approaches are quick, but can miss risk and struggle to represent masonry’s highly nonlinear behavior. This paper shows a proof-of-concept pathway to bring physics-based, nonlinear modeling into a workflow that can start from readily available façade imagery, offering a route to more informative prioritization without requiring costly, invasive data collection everywhere. The difference it could make is practical: more defensible triage for inspections and retrofits, and, if paired with publicly accessible street imagery, an enabling step toward near real-time, cross-border seismic risk mapping where formal surveys are hardest to conduct.

Perspectives

From a public-safety and equity-first policy perspective, this can be read as a critique of how seismic risk is managed: when monitoring networks are expensive and unevenly deployed, and building standards are poorly enforced, whole neighborhoods effectively become unseen until disaster strikes. In that framing, who gets measured becomes a political question, because measurement capacity influences which risks are acknowledged, funded, and retrofitted. The politically charged implication is that scalable, image-based screening (even if imperfect) can shift power toward accountability and preventive investment, by making it harder for authorities to claim uncertainty as an excuse for inaction. The paper positions its approach as predictive and potentially scalable via publicly accessible urban imagery toward real-time, cross-border risk maps, exactly the kind of visibility that can pressure institutions to prioritize retrofits and enforcement rather than waiting for post-disaster reconstruction cycles. It also acknowledges the core trade-off: broader accessibility comes at the cost of missing interior structural details, which is a reminder that cheap scale should be used to triage and direct resources, not to justify austerity or replace proper engineering follow-up.

mayar ariss
Massachusetts Institute of Technology

Read the Original

This page is a summary of: Seismic assessment of unreinforced masonry façades from images using macroelement-based modeling, Communications Engineering, August 2025, Springer Science + Business Media,
DOI: 10.1038/s44172-025-00487-2.
You can read the full text:

Read

Resources

Contributors

The following have contributed to this page