<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/a42174870ac7454cbb662cad3761ab76&quot; frameborder=&quot;0&quot; width=&quot;1920&quot; height=&quot;1440&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1440</height><width>1920</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1440</thumbnail_height><thumbnail_width>1920</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/a42174870ac7454cbb662cad3761ab76-4783c62268ef80f1.gif</thumbnail_url><duration>203.733</duration><title>Fire risk forecasting and optimal asset protection</title><description>This Loom presents a Hack Barner 2026 submission addressing DeepFire’s challenge of protecting the right assets and guiding response locations during an active wildfire. The team uses a two part approach: predicting where a fire will be after one and two hours, then combining that forecast with the value of assets across Catalonia to minimize damage. They integrate DeepFire’s fire prediction endpoint while also implementing and region calibrating the underlying open source model to improve data quality, with results not yet benchmarked on historical performance. The output is a dashboard with a local fire map, scenario specific forecast timelines, asset risk drilldowns, and an incident briefing in plain text for first responders.</description></oembed>