Virtual “Disaster Cities” to Provide Real-Life Resilience

After a massive earthquake struck Japan on March 11, 2011, teachers at Ukedo Elementary School ordered a swift evacuation to higher ground, saving the entire student body from the deadly tsunami that followed. Tragically, students at Okawa Elementary School remained in the schoolyard instead of moving to a nearby hillside, resulting in the loss of 74 students and 10 teachers.

Given the enormous impact of such split-second decisions, how can we better prepare people to make informed choices when facing an impending disaster?

Computer Science Professor Nalini Venkatasubramanian is exploring this question through a partnership between UC Irvine and Tohoku University in Japan. “After going through the 2011 earthquake and tsunami, researchers at Tohoku started conducting tsunami inundation modeling, so they have a huge amount of expertise in this type of planning,” says Venkatasubramanian. She and a team of researchers in UCI’s Donald Bren School of Information and Computer Sciences (ICS) have their own expertise in leveraging data gathered through the Internet of Things (IoT) to help protect vulnerable communities — particularly older adults.

Now, the two groups of researchers are collaborating on a three-year $1M grant funded through the National Science Foundation (NSF) and Japan Science and Technology (JST) agency. The joint NSF-JST project, Enabling Human-Centered Digital Twins for Community Resilience, aims to apply the concept of “digital twins” to disaster science, better analyzing the impact of natural disasters on humans and outlining real-time pathways to safety.

Protecting Older Adults

“The situation becomes even more dire when it comes to populations with specialized needs, like older adults or people with disabilities,” says Venkatasubramanian. “So, we’re focusing on leveraging data from these ‘digital twins’ to reduce the vulnerabilities of older adults during a natural disaster.”

The U.S. team includes Computer Science Professor Sharad Mehrotra and other ICS faculty; remote sensing and geographic information system (GIS) experts from Boston University; and experts from ImageCat Inc., a risk management innovation company in Long Beach. The researchers are also engaging with Julie Rousseau and Lisa Gibbs, geriatric medical experts from the UCI School of Medicine and Nursing.

“With medical multi-complexity, older adults face higher morbidity and mortality during disasters around the world,” says Rousseau. “With rapidly rising 65+ populations, Japan and the U.S. need innovative solutions to build equitable disaster resilience for vulnerable older adults with physical and/or cognitive disabilities.”

The project involves designing tools to create Disaster Digital Twin (DDT) cities that are representative of Sendai in Japan and Southern California in the U.S. These “cyberworld replicas” will highlight both what these communities have in common, as well as certain differences.

“While Southern California has a lot of continuing care facilities, housing larger groups of older adults together, the community structure is very different in Japan, where older adults are oftentimes more integrated into the general community, living independently,” says Venkatasubramanian. “Also, Southern California is more likely to experience a wildfire than a tsunami.” These types of differences will be considered during crisis modeling.

Building Human-Centered Digital Twins

“Researchers usually predict a disaster’s impact in a very mathematical sense,” says Venkatasubramanian. For example, she says they might provide a mathematical modeling of a tsunami’s inundation levels, indicative of the height and speed of the incoming water flow. “But then how,” she asks, “do we assess the impact on the broader community?”

The goal is to create a “virtual disaster city” that can not only map out the effects of a disaster but also analyze the impact on specific structures and the people within those structures — including data on their health and mobility needs.

“As we saw at Ukedo Elementary School, everyone there could move to higher ground for safety,” explains Venkatasubramanian. “But what if a main bridge had collapsed? What if you were home alone in a wheelchair? What if you were in a skilled nursing facility, hooked up to a ventilator? These are the kinds of questions our virtual cities hope to answer.”

The team is working with Japanese and California agencies with expertise in disaster resilience and aging populations, such as the International Research Institute on Disaster Services (IRIDeS) in Japan. Stakeholder workshops will enable the team to understand the older adult disaster landscape, leading to research that

• integrates diverse geospatial and human-centric data sources into a novel information architecture to enrich and align multiresolution spatiotemporal data;
• executes physics-driven hazard simulations; and
• simulates disaster processes and consequences in virtual disaster cities, focusing on an older population.

“Modeling communities and their interactions and needs as a digital twin for disaster response is a critical application with transformative impact for social good,” says Mehrotra. “Such modeling will include dealing with data incompleteness and uncertainty, finding methods for timely data collection and planning, and centering the required human elements — advancing ‘ditial twin’ technology in new and important directions.”

Furthermore, the team will build on lessons learned from the CareDEX project, another effort Venkatasubramanian is leading. CareDEX is a smart-space platform for securely exchanging customized care information between first responders, caregivers in senior housing facilities, and older adults.

Exploring Interdisciplinary, Data-Driven Solutions

Venkatasubramanian says that the larger goal of the Digital Twins project is to bring together a variety of stakeholders in exploring data-driven solutions. “You have geologists, tsunami experts, engineers, AI specialists and computer scientists, professional caregivers, government officials,” she says, “all looking at the sociological and societal contexts as we explore ways to build better technology solutions.”

She touched on this effort in February when giving a keynote talk in Sendai for the International Conference on Big Data for Disaster Response and Management in Asia and the Pacific. Sponsored by the Asian Development Bank, which invests in building climate resilience across more than 30 countries, the talk focused on emerging technologies for enabling resilient cyber-human infrastructure.

Such infrastructure could support a variety of populations, across the globe, experiencing a range of disasters, from floods to wildfires.

“Although the Digital Twins project is looking at a very specific context, we’re also considering what solutions translate across the international community and plan to expand our findings more globally,” says Venkatasubramanian. “We need to understand the structure of the community and its policies to simulate a variety of different scenarios to help people better prepare.”

— Shani Murray