Many countries are subject to severe natural disasters, and the effects can have devasting impacts, like destroying up to 200% of gross domestic product and stunting long-term economic growth. One way in which governments can help stem the financial impact is by increasing building resilience through planned maintenance and retrofits and strategizing investment of public funds, but governments struggle to collect and analyze enough data on public buildings to guide and prioritize maintenance tasks.
Gonzalo Pita, associate research scientist in the Whiting School of Engineering’s Department of Civil and Systems Engineering, and his team of researchers have developed a building condition assessment tool to support government efforts in ensuring structures are in sound condition and prepared to withstand natural disasters before they happen.
Published in Reliability Engineering & System Safety, the team’s assessment model has the potential to protect a country’s long-term fiscal stability.
“Having knowledge of a building’s physical condition allows governments to take proactive or retrofit measures if they’re needed, but assessing a building’s physical state requires in-person visits and hours of work conducted by specialist structural engineers, which is expensive, even for relatively simple buildings,” said Pita. “Machine learning techniques have helped populate a lot of missing data, such as footprint, roof type, and number of stories, but physical condition is more difficult to infer because it’s engineering-based and changes over time.”
The team’s new model gives public—and also private sector—property owners a flexible way to assess building conditions to help save time and resources. Their methodology uses mathematical approaches that separate a building into a hierarchy of component and whole-structure levels where deterioration naturally occurs and estimates degradation at each level over a given period of time.
The model does not take into account maintenance or reconstruction schedules, nor does it account for damage following a natural disaster; rather, its strength lies in capturing complex deterioration dynamics.
While several methods of building condition assessment exist, the researchers say that they are primarily empirical and are unable to provide analysts with the modeling detail needed to understand progressive deterioration for individual components.
Pita says that the model is intended to obtain condition estimates on an inventory of buildings to ultimately identify structures that need more detailed assessments. The model’s results provide a valuable tool for disaster preparedness, finance management, risk management, and preventative maintenance and planning.
“Our condition assessment model can help governments save public resources and money, especially in smaller, developing countries,” said Pita. “It’s a way to help ministries of finance and public works foresee future financial needs and better prepare for severe natural events.”
The study was prompted by Pita’s research and collaboration with countries to develop risk assessments, catastrophe models, and disaster risk financing. He became familiar with government processes, budget flows, and fiscal stability dynamics in small and developing nations. He anticipates that the model will help resource-constrained governments plan more efficiently through targeted public investments and protect fiscal stability and public assets.
With an eye towards the future, the team is already working on their next project: aiming to identify optimal interventions for public building inventories.
Study co-authors include Johns Hopkins University doctoral student Shaochong Xu and World Bank consultant Francisco Michati.