Building a more resilient city of angels (part 1)

In early January, Governor Gavin Newsom said: “November, December, now January − there’s no fire season, it’s fire year. It’s year-round.”

Facing the uncertainties of droughts, wildfires, blackouts, poor air quality, mudslides and flooding, here are some ideas from faculty at the USC Viterbi School of Engineering to build a more resilient City of Angels and beyond. 

“3D Printing Fireproof Homes”

Berok Khoshnevis, Louise L. Dunn Endowed Professor in Engineering and Distinguished Professor of Civil and Environmental Engineering, Astronautics and Aerospace and Mechanical Engineering

“Instead of new wooden buildings, the city should consider 3D printing new houses made of concrete. This would not only have greater resistance against fire, but increase the speed, safety and flexibility of re-construction.

My company, Contour Crafting, has a small fleet of three 3D printing machines in our possession, and we can build about 20 more machines over the next year.

In practice, we could start with the three machines, and, as each newly built machine is released by our factory, we will add it to the fleet of working machines. By the end of the year, 20-25 machines would be working concurrently. Currently, 3D printing only builds the concrete shell of the building.

To get an idea about the construction speed, note that a 2000-square foot house shell may be printed in two days. Using any other method, the process would be significantly slower.

Concrete is a non-combusting material; therefore, it never catches fire. It is also proven to have a high degree of resistance against fire created by combustible materials. Fire over long duration and extreme heat can still affect structural integrity of concrete by causing spalling or other damages. In most applications, however, concrete can be characterized as virtually fireproof.

I think what will be most essential to help those who have lost their homes to wildfire is first a quick accessory dwelling unit (ADU), built according to specific size and design selected from a variety of options offered in a catalog of choices. Construction 3D printing is an ideal choice for responding to this emergency construction which would rapidly provide shelter to the displaced homeowners. Once the owners are situated on their land, then bigger building can be constructed by 3D printing or other methods of owner’s choice.

As for the cost of 3D printing, the economy of scale would greatly help to reduce the cost of special 3D printable material, which is currently higher than ordinary concrete. Also, the current scarcity problem of technical operators of construction 3D printers will be alleviated if many such operators are trained for building many buildings. Currently, the number of buildings printed by 3D printing is too few for the workforce to consider getting training and entering the profession.”

“Avoid Powerlines in Wildfire-Prone Lands. Generate Energy Where You Live.”

Kelly Sanders, Dr. Teh Fu Yen Early Career Chair and Associate Professor of Civil and Environmental Engineering

“The key to any fire is a spark — and our power system has become a common source of recent ignitions, with power lines spanning long distances through wildfire-prone lands to bring electricity from power plants to population centers.

While it is easy to point fingers at who should be held responsible for malfunctioning infrastructure, the hard truth is that electricity ratepayers cannot afford to fund a fireproof electricity system. So the question becomes: how do we build a power system that minimizes fire risk while maximizing carbon free energy generation that slows the threat of climate change?

Part of the answer for California is to invest more in local solutions: generating electricity closer to where consumers live and using our existing electricity system more efficiently. Solar panels are a good start, but they must be paired with batteries, and other distributed energy resources that can provide and/or store electricity at night. We also need to look at next-generation clean energy resources like offshore wind turbines, clean hydrogen generated with excess renewables, and advanced nuclear microreactors.

While some of these technologies are ready for implementation, others will need nudges to develop viable markets. However, we won’t solve this through local power generation alone. Every unit of electricity we don’t use is one that doesn’t need to flow through high fire-risk regions. We need to maximize energy efficiency measures, invest in high-density housing and microgrids to reduce the energy intensity of our lifestyles and be more resilient to disruption, develop financial incentives that encourage people to use electricity when it is abundant, and similarly, integrate technologies that can help optimize energy flows through our homes, buildings, distributed energy resources, and batteries to match power system dynamics.

Then we can begin retiring the highest-risk portions of our power systems while strategically investing in state-of-the-art fire mitigation strategies for the parts of the transmission system that remain critical.”

“AI Powered Buildings That Adapt to the Unknown”

Burçin Becerik-Gerber, Dean’s Professor and chair of the Sonny Astani Department of Civil and Environmental Engineering

“I’m interested in how the buildings that we occupy right now impact our health and well-being. I’m not just talking about physical health, but also mental health. It’s a new and emerging field in civil and environmental engineering. We call this human-building interaction.

Imagine AI-powered buildings that can ‘sense’ and ‘adapt dynamically’ to changing conditions, both indoors and outdoors.

For example, if the outdoor air quality metrics are poor, are there ways to trigger some of the building systems to improve the air quality? This dynamic response would involve adjusting indoor systems like ventilation and filtration based on outdoor environmental conditions (e.g., during wildfires when air quality deteriorates).

Also, buildings might use sensors in the water faucets to detect contamination after wildfires, or smart systems [could] kick in and clean the air or clean the water when environmental hazards emerge.

I’m also imagining protective layers that lay on buildings to protect them, such as automated systems that shield buildings from fires or other environmental hazards. Think of the suits firefighters are wearing — can we have suits for our buildings to protect them? While still speculative, such systems could potentially save lives and reduce damage during disasters.”

“Bad Air, Better Clarity”

Jiachen Zhang, Assistant Professor of Civil and Environmental Engineering and Spatial Sciences

“Following the recent wildfires, I’ve received numerous inquiries about air quality concerns. While there are many online resources and reports available, gathering and interpreting this information can be overwhelming.

Much of the current guidance remains qualitative rather than quantitative. For instance, while we know that emissions from burning structures tend to be more toxic than those from forest fires, quantifying these risks requires detailed chemical analyses, measurements and public health studies. Additionally, the Air Quality Index (AQI) doesn’t capture the complete picture — it may overlook certain pollutants, such as volatile organic compounds (VOCs) or the varying toxicity of different particulate matter (PM) species. As a result, even when AQI levels appear acceptable, potential health risks may still exist.

To mitigate wildfire-related air quality issues, the simplest immediate actions include using an air purifier with a HEPA filter and activated carbon and wearing masks when outdoors.

However, uncertainties remain about when the air is genuinely safe to breathe. For example, wind can re-suspend particles from burned structures, prolonging exposure risks. This highlights the need for more comprehensive air quality measurements, quantitative chemical analyses, and epidemiological studies to better inform public health guidelines during and after wildfires. As someone who lives close to the Eaton fires, I would also like to invite researchers to conduct air quality sampling in my backyard to support these efforts. I think we need interdisciplinary efforts to address wildfire risk.”

“AI to Anticipate Extreme Fires”

Assad Oberai, Hughes Professor and Professor of Aerospace and Mechanical Engineering

“One remarkable aspect about the L.A. fires was the confluence of several very rare factors that made them especially disruptive. This included a long stretch of very little or no precipitation in the region, a very strong Santa Ana wind event with gusts up to 90 mph, and thick vegetation (fuel) generated by good rains in the previous year. These three ‘outliers’ have contributed to the ferocity of the wildfires and made the firefighting response especially challenging.

However, some of this could have been anticipated. We have at our disposal physics-based and data-driven tools that can simulate the progression of wildfires under a wide range of fuel, wind, and climatic conditions. With some additional work, one could use these tools to simulate what the progression of ‘virtual wildfires’ would look like even under these circumstances that are rare. This could provide us with a measure of the intensity and scale of these fires before they occur and perhaps make us better prepared to prevent and fight them.”

Barath Raghavan with banana and fire illustrations
Using computer modeling, Barath Raghavan found that banana orchards could help prevent wildfire spread. (Photo/Noé Montes)

“Bananas to the Rescue!”

Barath Raghavan, Associate Professor of Computer Science and Electrical and Computer Engineering

“One innovative approach to wildfire resilience is integrating food production with fire mitigation strategies. For example, planting fire-resistant crops like banana trees can act as natural firebreaks.

Their high-water content and low flammability make them effective at slowing or even stopping fires spreading into high-risk neighborhoods near the wildland-urban interface. Beyond fire prevention, this could offer added economic and environmental benefits like boosting local food production, community involvement and sustainable land use practices.

Another promising avenue we’ve been exploring is the use of artificial intelligence and automated crowdsourcing for wildfire detection and response. In a recent study, we found that a network of low-cost mobile phones mounted on properties in high fire threat areas could help identify fires up to 3,000 feet away and successfully map wilderness blazes within 180 feet of their origin.

By combining ecological solutions like fire-resistant crops and new technology like AI, I believe we can tackle wildfire risks on several fronts. Strengthening community resilience by engaging residents as active participants in wildfire detection and employing innovative land management practices are crucial in making Los Angeles more sustainable and resilient in the face of future wildfire threats.”

“The Hills Are Alive…With Autonomous Sensors and Drones”

Mahta Moghaddam, Ming Hsieh Chair in Electrical and Computer Engineering and Distinguished Professor of Electrical and Computer Engineering

“What could have helped both the Palisades and the Eaton fire responses, and importantly, what will help to address similar circumstances (or situations) in the future, is if autonomous systems were in place.

Such systems would have (1) vigilantly observed (based on the risk factors that we knew existed): high winds, dryness, hillside geometry, etc.), (2) immediately detected the first ignition, (3) immediately thereafter generated autonomous decisions on what to do based on various algorithms (including machine learning and AI schemes), and (4) immediately thereafter launched autonomous drones to put out the first small fires.

We need to create autonomous systems with humans-in-the-loop for sanity checks but also that perform with beyond-human speed (or transcend human speed). There are many technologies that need to be developed for all of the above steps. We need to develop observational technologies that are easy and scalable to deploy and can “wake up” on applicable cues. We need to develop closed-loop reliable autonomous decision making schemes. And we need to develop resilient drone/UAV technologies that can fly in harsh conditions. The latter may not be all that we need, but I think they will be extremely helpful in controlling wildfires at early stages and before things get out of hand.”

“Using AI for Controlled, Preventive Burns”

Yolanda Gil. Research Professor of Computer Science and Spatial Sciences and Principal Scientist at the USC Information Sciences Institute

“We have used AI for preventive fires in a five-year NSF-funded project led by UCSD.  A preventive fire is controlled by a fire manager and targets a site where vegetation has grown too much and has the potential to lead to a mega-fire.  A lot more preventive fires are needed in order to prevent mega-fires, but they are expensive and hard to do at scale because there are very few fire managers who do this by hand and without using scientific fire models.

Our project focused on using AI to help fire managers decide how to plan the controlled fires.  Once a site is identified as problematic, our research has been on selecting appropriate data and models for the area, then running simulations to help with decisions.  The simulations help understand tradeoffs in the design of controlled fires, such as when to do the fire depending on wind and weather patterns, what percentage of vegetation to burn (since leaving some vegetation can help with rain and mudslides), and also what level of air quality is acceptable depending on burning faster or slower.

Lots of challenges here, for example utility companies would be responsible for a lot of these preventive fires, but when they are not done there is often no consequence, ie no terrible wildfires, so their incentive is very limited.”

“A Virtual Ally in Your Pocket”

Sharon Mozgai, Director, Virtual Human Therapeutics Lab, USC Institute of Creative Technologies

“Los Angeles faces critical challenges in recovery from fires, including trauma and grief, overwhelmed mental health services, limited access to emergency resources, and inequitable distribution, particularly affecting marginalized communities. Battle Buddy, a virtual human-led AI system delivered via mobile phone application, could be tailored to address these needs and bolster L.A.’s resilience.

As a compassionate, trauma-informed interface, the virtual human component could support victims experiencing PTSD and anxiety, while the underlying AI agents could enable real-time analysis of needs, scalable resource allocation and personalized guidance to essential services like shelters, medical aid, food distribution, legal assistance and rebuilding support.

Together, these innovations could fill resource gaps, support equitable access, and strengthen community recovery efforts.”

“Investigation Before Condemnation”

Najmedin Meshkati, Professor of Civil and Environmental Engineering, Industrial Systems Engineering, and International Relations

“I have served as a member or adviser to national- and state-level investigations of events including gas leaks, oil spills, nuclear reactor accidents, refinery explosions and, most recently, aviation mishaps.

In my view, the Los Angeles fires call for a similar investigation that is technically sound, multidisciplinary, unbiased, apolitical and independent. U.S. Sen. Adam Schiff of California has called for convening such a review.

To quote a saying often attributed to Albert Einstein: ‘Condemnation without investigation is the height of ignorance.’

Natural disasters such as wildfires, earthquakes and tsunamis often serve as triggers. Devastating on their own, these events can have far more catastrophic aftermaths that are shaped by human choices. Nature delivers the initial blow, but a complex interplay of human, organizational and technological factors can either mitigate or worsen the consequences.

I believe human operators and first responders constitute society’s first and the very last layer of defense against death and destruction in the crucial moments following natural disasters and technological systems failures – serving as our immediate shield, intermediate mitigator and ultimate savior.

I saw this when I served on a National Academy of Sciences committee that studied the 2011 Fukushima nuclear accident in Japan. The explosions and radioactive releases at the Fukushima Daiichi plant were triggered by an earthquake and tsunami, but a Japanese high-level review concluded that this event was a “manmade disaster” – one born of human and organizational failure at the utility and governmental levels.

The fate of the Onagawa Nuclear Power Station, just 39 miles from Fukushima, was also notable. Although Onagawa was closer to the earthquake’s epicenter and faced an even more powerful tsunami, the reactors there – which were identical in type and age to Fukushima’s and subject to the same regulations – emerged almost unscathed. This stark difference demolished any argument that Fukushima’s failure was inevitable, an act of God or purely nature’s fault.

The most straightforward way to set up a high-level review of the Los Angeles wildfires would be for the Trump administration and Congress to direct the National Academies of Sciences, Engineering, and Medicine and the National Research Council to establish an independent commission.

The National Academies have a reputation for producing independent, rigorous and nonpartisan studies. They screen members thoroughly for technical expertise and conflicts of interest. All of their studies go through formal peer review, which helps ensure that they are scientifically accurate and credible.

I see the Los Angeles fires as a stark warning to communities nationwide. There is a widening gap between intensifying climate-induced extreme events that are becoming Earth’s new normal, and municipal planning, preparedness and response capabilities.

Meeting these unprecedented challenges requires a paradigm shift in public policy. To protect public safety, officials and planners will have to proactively confront scenarios that may recently have seemed unthinkable.”