# AI Project Could Help Cities Find People Left Behind After Disasters > When disaster hits—a flood, an earthquake, or something else that's hard to imagine—emergency teams have to move fast and make tough choices. Where’s the worst ![AI Project Could Help Cities Find People Left Behind After Disasters](/media/2026/09/Positivity---2026-09-24T103515-363-591a6930.webp) ## When disaster hits—a flood, an earthquake, or something else that's hard to imagine—emergency teams have to move fast and make tough choices. Where’s the worst damage? Which roads are open? How many families need help? Who's already gotten assistance? And maybe the hardest one to answer: > Who hasn’t been reached yet? That last question is tricky for all sorts of reasons. In the chaos, phone lines go down, cell towers get overwhelmed, and information comes in from all angles—calls, text messages, community leaders, emergency workers, people on social media, all in different formats and languages. Some folks can’t contact anyone at all. Others live in neighborhoods you won’t find on any official map. Now, there’s a new project called TRACE, powered by AI, that’s meant to tackle this head-on. It grabbed first place at the 2026 AI x City Climate Action Hackathon, chosen over dozens of ideas from around the world. The idea behind TRACE is simple but kind of brilliant: use AI to make sense of all those conversations with affected communities, turning scattered reports into clear information so cities can figure out not just who needs help, but who’s still missing from the radar. > That can save precious time when every hour counts. ## Finding the Hardest People to Reach Picture a city under water. Emergency phones ring off the hook. One neighborhood says families are trapped. Another mentions stranded elderly folks. A community group reports in a language that’s not the official one. A volunteer says the power’s out for blocks. Meanwhile, someone in another neighborhood can’t get a call out at all and doesn’t exist in the database. All the pieces are out there, but they’re scattered. It’s basically a giant puzzle, and missing a piece means someone could get overlooked. TRACE is designed to help with exactly that. According to the Global Covenant of Mayors, TRACE takes these messy, real-world conversations—even in regional languages like Zulu—and pulls out the details that cities need. It finds affected families and gets a feel for what they need right now. But here’s the really interesting part: > TRACE doesn’t just help cities see who’s on the list. It helps them spot who isn’t. ## How TRACE Turns Talk Into Action People don’t talk in tidy categories, especially in an emergency. They say things like: - “My mother’s stuck at home and can’t walk.” - “The water’s up to the second floor.” - “We have enough food for now, but who knows about tomorrow?” - “Our road’s blocked.” - “My neighbor hasn’t been seen since yesterday.” - “We need a safe spot for the kids.” There’s valuable info in all that. But someone has to hear it, make sense of it, organize it, and compare it to everything else that’s coming in. That’s where AI—and TRACE in particular—comes in. It listens to recorded conversations and turns scattered statements into structured information that helps teams on the ground. So, instead of people wading through every single report, the system highlights needs, patterns, and which households are affected. Response teams spend less time buried in paperwork and more time out there helping. ## Knowing Isn’t Enough—You Have to Actually Reach People In an emergency, there’s a difference between knowing something’s happened, knowing who’s affected, and knowing who’s actually gotten help. Let’s say you know there are a thousand families in a flood zone. Great, but what if teams have only been able to contact 700? What about the other 300? Without full information, they stay invisible. TRACE helps shine a light on those blind spots. The tech is built to help cities find people left out of emergency maps—so support can get to the folks everyone else might miss. > That’s the real point: not just making charts or dashboards, but answering the most human question in a crisis—who still needs us? ## Why Local Languages Matter Disasters don’t only hit cities where everyone speaks the same official language. Communities are often multilingual. At home, people talk in whatever language they feel most comfortable using, especially when they’re scared or stressed. Systems that only handle formal, standardized reports miss a lot here. TRACE was designed to work with languages like Zulu for exactly that reason. If your tech only understands perfect formal reports, you miss all sorts of critical info. But if it can process different languages and dialects, then suddenly a lot more people are heard. ## A Solution Born in a Global Challenge TRACE came out of the 2026 AI x City Climate Action Hackathon, which drew nearly 700 people from 77 countries. Out of 75 projects, 11 teams—representing countries like the UK, Colombia, Kenya, India, and the US—made it to the finals. Different cities face different emergencies—floods in one place, heatwaves or fires in another. But the core challenge stays the same: > Can cities figure out what’s really happening, fast enough to help people? The hackathon gave teams the chance to attack that question with AI and similar technologies. ## TRACE Puts Vulnerable People First This project isn’t just about collecting data. It’s focused on understanding who is affected by a disaster and what they truly need. That’s important because disasters hit people differently. Some can’t evacuate on their own. Elderly residents, people without reliable transportation, those in informal settlements, folks who don’t speak the dominant language, or families who can’t get online—they’re the hardest to find, and often the most in need. So, emergency response can’t stop at just knowing there’s a flood. Responders need to know about real people: where they are, how they communicate, and why they might be missed. ## AI: Another Tool in the Emergency Kit TRACE is part of a wave of AI being used for disaster relief. There are tools that analyze satellite photos to spot damaged buildings. Other models monitor changes in roads or population movement. Weather models get sharper, and now language-based AI can pull meaning from the flood of reports and conversations. Microsoft’s AI for Good program, for example, uses satellite data to gauge damage, and its HASTE platform helps with route planning during disasters. The UN Global Pulse project has a system, DISHA, that examines anonymous mobile data to guess at population changes post-disaster. > The point isn’t that these tools replace people. They help people get a clearer, faster read on a fast-changing situation. ## Finding People Is Just Step One Identifying where affected people are doesn’t solve their problems. One family needs clean water. Another wants shelter. A neighborhood is cut off until a road is cleared. Information is only valuable if it leads to action—if responders can figure out not just who still needs help, but why help hasn't arrived. Sometimes the reason is simple: a blocked road, a full shelter, or a community using a system the authorities weren’t monitoring. Better information helps emergency teams figure out what’s slipping through the cracks. ## The Human Side Still Counts Most It’s easy to get starry-eyed about what AI can do—but it can’t solve everything. Human judgment makes the calls, not the algorithm. Someone has to decide if it’s safe to send a team in. A community leader might be the only one who knows which homes have elderly people. And nothing replaces the trust and local knowledge of actual people on the ground. TRACE is built with that in mind. The plan is to push from hackathon demo to real-world city trial and give it a go under messy, stressful, real-life conditions. ## From Idea to City Trial Winning a hackathon is one thing. Emergency situations are something else. When phone networks are spotty and everyone’s stressed, any new tool has to fit with the stuff cities already depend on. That’s why TRACE’s city pilot matters: it’ll team up with a real city, join the Build for Earth program, and see what works or fails in practice. This is where the real tests happen: - Can first responders use it easily? - Can it handle variety in languages and accents? - Is the info accurate? - Does it reliably find the people everyone else missed? - Can it keep sensitive data safe? City pilots aren’t just a nice to have—they’re critical. ## Privacy and Trust Aren’t Optional You can’t run a system like TRACE without getting privacy right. Conversations can be deeply personal. Descriptions of health, family situations, or exact locations all need to be protected. And in a disaster, people have to believe the info they share will help, not hurt, them. > So, as TRACE moves into the field, nailing down privacy rules and building trust will matter as much as the AI’s accuracy. ## A Future Where No Neighborhood is Invisible One of the most hopeful ideas behind TRACE is simple: making hidden people visible. Cities have maps and databases and plans. But disasters don’t wait for the right moment, and they change things fast. Bridges wash out, power fails, entire neighborhoods get cut off. The map from the morning might be useless by noon. If AI can keep updating what’s known, emergency teams can move faster and more accurately. ## Letting Communities Speak for Themselves Here’s another angle: TRACE doesn’t just collect data for the city, it also gives communities a way to say what they see and know. People on-site can report which streets are impassable or which neighbors need extra help, details someone watching from a remote command center might never notice. Their voices add context and texture to the overall response. ## A Global Problem Needs Many Kinds of Solutions TRACE was just one of 75 entries in the hackathon. Others tackled problems like extreme heat, misinformation, waste management, and early warnings. No single tool fixes everything. Some cities need faster alerts; others need ways to find vulnerable groups or better connections between residents and emergency teams. Cities are different, and so are the solutions. ## Turning Information into Action During a disaster, information gets old fast. A road that was open an hour ago might be underwater now. A family that needed food yesterday needs rescue today. > Quick and accurate information can be the difference between confusion and fast help. TRACE’s promise is to help cities listen, understand, spot gaps, and reach people quickly. ## A Small Project With a Big Human Goal “Artificial intelligence” can sound cold, but TRACE has a clear, human mission: - Find people who need help. - Understand what they need. - Notice who’s missing from the response. - And then help cities actually reach them. Behind every bit of data is a person waiting—an older resident who can’t get out, a parent searching for safety for their kids, a whole community waiting for water or news, a neighbor hoping someone knows they’re there. Better information can mean a lot. ## Technology That Helps People Find Each Other Disasters cut people off. Routines vanish. Families are separated, and it’s easy to lose contact, even with those who want to help. Technology can’t fix everything, but used right, it reconnects communities. A satellite might show which roads are gone. Phone data hints at where people moved. AI organizes reports, spots what’s missing, and highlights those who haven’t been reached. > None of the tools tells the full story. But taken together, they give emergency teams a sharper, more connected picture. ## Building Cities That Leave Fewer People Behind Extreme weather is putting pressure on cities everywhere—floods, heat, storms, you name it. Being resilient isn’t just about better buildings or smarter drains; it’s about knowing who’s in your city, who’s most vulnerable, and making sure you can reach them when things fall apart. TRACE is still young, and its city pilot will show what it’s really made of. But its approach—to use AI to connect people, not just predict disasters—is the right track. Because a resilient city isn’t just one that can spot a flood coming. It’s one that can answer: > “Who is here? Who needs help? Who have we reached? And who’s still waiting?” If technology can help cities answer those questions more quickly and clearly, emergency response gets more human, not less. That’s the promise TRACE puts on the table—a future where, when the worst happens, fewer people fall through the cracks, and more are found, heard, and helped. Explore Related Topics Take the Positivity Quiz More in Good News Share this article ### Stay Inspired Get a daily dose of positivity delivered to your inbox. ### Discover More on Positivity What Is Positivity? Affirmations Quotes Meditations Psychology 🛠️ Tools Today's Inspiration ## You Might Also Enjoy ![Hong Kong Students Are Designing Greener Spaces for Their Schools](/media/2026/09/Positivity---2026-09-22T125512-543-604cdd0e.webp) Good News ## Hong Kong Students Are Designing Greener Spaces for Their Schools ![Borneo Villages Are Earning More by Protecting Their Forests](/media/2026/09/Positivity---2026-09-22T124306-749-7366aebc.webp) Good News ## Borneo Villages Are Earning More by Protecting Their Forests ![Singapore Launches New Program to Help Students Become Future Leaders](/media/2026/09/Positivity---2026-09-21T115319-952-f30bbdb4.webp) Good News ## Singapore Launches New Program to Help Students Become Future Leaders ![33 Water Utilities Around the World Are Turning Waste Into Resources](/media/2026/09/Positivity---2026-09-18T110151-502-5f588f6e.webp) Good News ## 33 Water Utilities Around the World Are Turning Waste Into Resources ![Australia Opens New Hub for the Future of Aviation](/media/2026/09/Positivity---2026-09-17T131816-959-50a2cdee.webp) Good News ## Australia Opens New Hub for the Future of Aviation ![Berlin Gives Renters Stronger Protection From Rising Rents](/media/2026/09/Positivity---2026-09-15T113448-418-23e86faf.webp) Good News ## Berlin Gives Renters Stronger Protection From Rising Rents ## Leave a comment