US Public Data LabU.S. public data, made readable
Data areas
Federal Register Congress bills CISA KEV / NVD SEC EDGAR USAspending Defense contracts Trade Economy Energy Places FDA recalls AI Research Grants NIH research grants Fiscal & national debt
Tools
Search Recent Rankings Compare Money flows Timeline Compare indicators Topics
Profiles
Company profiles Agency profiles Glossary Contact
EN▾
日本語 English
$2,000,000 TIP-CHIPS KTA-5 Disaster mgmt

NSF award $2M: AI for wildfire evacuation — built for roads where self-driving and human-driven vehicles mix (University of Southern California)

University of Southern California ・California ・Started Sep 2026

NSF awarded $2,000,000 to the University of Southern California for "FIRE-WUI: Agentic Artificial Intelligence for Wildfire Evacuation under Mixed Autonomy," addressing how an evacuation works on roads carrying both human-driven and autonomous vehicles. The program covers disaster management; the period runs September 2026 to August 2030.

Grant overview (primary data)

  • Award amount$2,000,000
  • RecipientUniversity of Southern California (California)
  • ProgramTIP-CHIPS KTA-5 Disaster mgmt
  • Period2026-09-01 〜 2030-08-31
  • FunderU.S. National Science Foundation (NSF) / NSF

Key points

  • NSF Award 2623936, "FIRE-WUI: Agentic Artificial Intelligence for Wildfire Evacuation under Mixed Autonomy," to the University of Southern California (CA).
  • Program: TIP-CHIPS KTA-5 Disaster mgmt. $2,000,000 obligated and estimated.
  • WUI is the Wildland-Urban Interface, where forest or open country meets built-up area — where wildfire damage concentrates.
  • Mixed autonomy means roads carrying both human-driven and autonomous vehicles.
  • Award date 2026-08-20; period 2026-09-01 to 2030-08-31. No research abstract is included in this data.
  • Evacuations may direct contraflow and shoulder running, so behaviour learned under normal conditions does not transfer as it stands.

1The place called WUI

WUI in FIRE-WUI stands for the Wildland-Urban Interface, the band where forest or open country meets built-up area. Wildfire damage concentrates there: fuel is abundant, and so are the people who have to leave and the roads they must leave by. The difficulty of evacuation lies less in the speed of the fire than in traffic converging on a limited number of roads until it stops moving.

2The premise of "mixed autonomy"

The condition set in the title is a road carrying human-driven and autonomous vehicles at once. That is a projection of the future and, in places, already the present. Mixing is hard because what the machine has to predict is not only other machines. In an evacuation the difficulty compounds: ordinary traffic rules change, and driving against the flow or along the shoulder may be directed.

How to move vehicles when learned normal-condition behavior does not apply is the question.

3"Agentic" AI

Agentic is used for AI that does not merely answer a prompt but judges and acts toward a goal. Directing an evacuation requires a continuous sequence of such judgments: which vehicles onto which routes, when to change the instruction. This data carries no abstract, however, so what will actually be built is not recorded, and this site does not write beyond what the title states.

4What was learned in normal conditions does not apply

The condition set by the title is human-driven and autonomous vehicles sharing the same road. Mixed autonomy is hard because what a machine has to predict is not only other machines. In an evacuation, a further element joins that difficulty.

The road in normal conditionsThe road during an evacuation
Traffic rules stay constantContraflow and shoulder running may be directed
Flow follows everyday demandTraffic converges on limited roads all at once
Learned behaviour transfers directlyBehaviour learned under normal conditions does not apply
Delay shows up as inconvenienceDelay bears directly on lives

Wildfire damage runs worst in the belt where wildland meets urban development. Alongside the abundance of fuel, the people leaving and the roads they use are concentrated there. The difficulty of evacuation lies less in the speed of the fire than in traffic converging on limited roads until it stops moving. This data carries no abstract, so what mechanism will be built is not recorded here.

Why it matters

Where AI helps in a disaster is less at the prediction stage than at the stage of moving things once a prediction exists. Traffic direction involves many parties issuing and receiving instructions under rules that differ from normal operation. Where autonomous vehicles are mixed in, reconciling on-vehicle decision-making with instructions from a control side becomes the implementation problem — a question for vehicle manufacturers and local governments alike.

FAQ

What is the WUI?
The Wildland-Urban Interface, the band where forest or open country meets built-up area. Fuel is abundant and people and roads are concentrated, so wildfire damage is heaviest there.
Why is mixed autonomy hard?
Because what the machine must predict is not only other machines. In an evacuation ordinary traffic rules also change, so learned normal-condition behavior stops applying.
What is agentic AI?
AI that does not merely answer a prompt but judges and acts toward a goal. What will actually be built here is not recorded in this data.

Related profiles

Related explainers

Sources (primary)

Source: NSF Award Search (U.S. National Science Foundation, public domain). Amounts are the obligated amount. For privacy, we do not handle principal investigator names.

#NSF#AI#Disaster response#Wildfire#Autonomous vehicles#Evacuation#Research infrastructure

← Back to the AI research grants ranking

Disclaimer: This site independently summarizes and classifies information based on official data sources. Always verify the latest and accurate information with the official sources. Content on finance, health, legal, and security is information, not advice. This site is not an official website of the U.S. government.