{"id":2986332,"date":"2026-10-30T09:00:38","date_gmt":"2026-10-30T16:00:38","guid":{"rendered":"https:\/\/www.esri.com\/arcgis-blog\/?post_type=blog&p=2986332"},"modified":"2026-10-30T10:47:26","modified_gmt":"2026-10-30T17:47:26","slug":"arcgis-for-microsoft-fabric-zombie-outbreak","status":"publish","type":"blog","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/maps-for-fabric\/analytics\/arcgis-for-microsoft-fabric-zombie-outbreak","title":{"rendered":"Patient Zero in Redlands – Tracking a Zombie Outbreak with ArcGIS for Microsoft Fabric"},"author":297282,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"open","ping_status":"closed","template":"","format":"standard","meta":{"_acf_changed":false,"_searchwp_excluded":""},"categories":[23341,22941],"tags":[779142,779152,780915,780882],"industry":[],"product":[780893],"class_list":["post-2986332","blog","type-blog","status-publish","format-standard","hentry","category-analytics","category-mapping","tag-arcgis-for-microsoft-fabric","tag-arcgis-geoanalytics-for-microsoft-fabric","tag-arcgis-maps-for-microsoft-fabric","tag-microsoft-fabric","product-maps-for-fabric"],"acf":{"short_description":"Zombies emerge in Redlands the day before Halloween. See how spatial analysis in ArcGIS and Microsoft Fabric tracks the spread.","flexible_content":[{"acf_fc_layout":"content","content":"

It started on New York Street…<\/h2>\n

On the morning of October 30th, fifty zombies shuffled out of 380 New York Street in Redlands, California. A number of questions quickly came up: by day 30 where have they spread across the Inland Empire? Which neighborhoods fell first? Which ones held out? and how fast the front moved each day?<\/p>\n

As you are probably aware, no one was harmed in the making of this blog post. The outbreak is a simulation built in Microsoft Fabric, using ArcGIS GeoAnalytics for Microsoft Fabric<\/a> for the spatial heavy lifting, ArcGIS Maps for Microsoft Fabric<\/a> to visualize, validate and verify the information, and I’ll propose some different options to enable decision-makers to understand the extent of the zombie infestation. It reads real US census data and information, models the spread across a grid of hexagons, and writes the results to OneLake and ArcGIS Online.<\/p>\n

It is a Halloween project, so it is meant to be fun. But by the end I hope you see it the way I do: a template for spatial problems your business actually has.<\/p>\n

 <\/p>\n"},{"acf_fc_layout":"image","image":{"ID":2986356,"id":2986356,"title":"Zombies_Day30","filename":"Zombies_Day30-scaled.png","filesize":5051964,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-scaled.png","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/maps-for-fabric\/analytics\/arcgis-for-microsoft-fabric-zombie-outbreak\/zombies_day30","alt":"Zombie extent","author":"297282","description":"","caption":"30 Days after the initial outbreak, we can see how things are spreading and the density of the infestation.","name":"zombies_day30","status":"inherit","uploaded_to":2986332,"date":"2026-10-28 10:57:30","modified":"2026-10-30 14:06:19","menu_order":0,"mime_type":"image\/png","type":"image","subtype":"png","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":2560,"height":1526,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-213x200.png","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-scaled.png","medium-width":438,"medium-height":261,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-scaled.png","medium_large-width":768,"medium_large-height":458,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-scaled.png","large-width":1812,"large-height":1080,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-1536x916.png","1536x1536-width":1536,"1536x1536-height":916,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-2048x1221.png","2048x2048-width":2048,"2048x2048-height":1221,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-780x465.png","card_image-width":780,"card_image-height":465,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_Day30-1812x1080.png","wide_image-width":1812,"wide_image-height":1080}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"content","content":"

Why zombies?<\/h2>\n

A zombie outbreak is a spatial diffusion problem wearing a scary costume. Something starts at a point, moves to nearby places, grows or shrinks depending on the local conditions, and leaves a footprint that changes through time and evolves.<\/p>\n

That is the same pattern as many workflows done in GIS: a disease moving through counties, a wildfire front, a power outage cascading across a grid, or a new product catching on store by store. Zombies just make it easier to talk about and see capabilities and how they may apply to your daily workflows.<\/p>\n

They also have a respectable research history. Munz and colleagues published an SZR (susceptible, zombie, removed) model<\/a> in 2009, and a Cornell team simulated a zombie outbreak<\/a> across the continental US in 2015. I borrowed the SZR idea and included geography.<\/p>\n"},{"acf_fc_layout":"content","content":"

The stack<\/h2>\n

All the analysis runs in one Fabric notebook. GeoAnalytics for Fabric adds more than 160 spatial functions to Spark, plus readers and writers for ArcGIS feature services, so no data has to be downloaded by hand.<\/p>\n

Public census and vulnerability data flow in from ArcGIS feature services; results land in OneLake and, optionally, ArcGIS Online<\/a>.<\/p>\n"},{"acf_fc_layout":"image","image":{"ID":2986350,"id":2986350,"title":"ZombieArchitecture","filename":"ZombieArchitecture.png","filesize":973932,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture.png","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/maps-for-fabric\/analytics\/arcgis-for-microsoft-fabric-zombie-outbreak\/zombiearchitecture","alt":"Conceptual Architecture","author":"297282","description":"","caption":"","name":"zombiearchitecture","status":"inherit","uploaded_to":2986332,"date":"2026-10-28 10:12:39","modified":"2026-10-28 10:13:04","menu_order":0,"mime_type":"image\/png","type":"image","subtype":"png","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":1536,"height":1024,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture-213x200.png","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture.png","medium-width":392,"medium-height":261,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture.png","medium_large-width":768,"medium_large-height":512,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture.png","large-width":1536,"large-height":1024,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture.png","1536x1536-width":1536,"1536x1536-height":1024,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture.png","2048x2048-width":1536,"2048x2048-height":1024,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture-698x465.png","card_image-width":698,"card_image-height":465,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/ZombieArchitecture.png","wide_image-width":1536,"wide_image-height":1024}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"content","content":"

ArcGIS Maps for Fabric<\/strong> closes the loop inside the workspace. It is a Fabric workload, that creates a new item like a notebook or a lakehouse, that can map lakehouse data directly. That means the data team can check each run on a real map before anything is published. The same map can also pull in authoritative content from your ArcGIS Online organization or from the ArcGIS Living Atlas<\/a>, so the outbreak can sit on top of roads, imagery, or your own hosted layers.<\/p>\n"},{"acf_fc_layout":"image","image":{"ID":2986362,"id":2986362,"title":"First Infested Day","filename":"Zombies_FirstInfestedDay-scaled.png","filesize":3194084,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-scaled.png","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/maps-for-fabric\/analytics\/arcgis-for-microsoft-fabric-zombie-outbreak\/zombies_firstinfestedday","alt":"First Infested Day","author":"297282","description":"","caption":"Understanding how things spread in a temporal basis is important. Here we see how the Zombie hoard has moved on a day by day basis after 30 days.","name":"zombies_firstinfestedday","status":"inherit","uploaded_to":2986332,"date":"2026-10-28 11:06:21","modified":"2026-10-30 14:08:00","menu_order":0,"mime_type":"image\/png","type":"image","subtype":"png","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":2560,"height":1526,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-213x200.png","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-scaled.png","medium-width":438,"medium-height":261,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-scaled.png","medium_large-width":768,"medium_large-height":458,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-scaled.png","large-width":1812,"large-height":1080,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-1536x916.png","1536x1536-width":1536,"1536x1536-height":916,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-2048x1221.png","2048x2048-width":2048,"2048x2048-height":1221,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-780x465.png","card_image-width":780,"card_image-height":465,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2026\/10\/Zombies_FirstInfestedDay-1812x1080.png","wide_image-width":1812,"wide_image-height":1080}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"content","content":"

Building the world<\/h2>\n

Before anything can spread, the notebook needs a map of who lives where. It builds that in three steps.<\/p>\n

1. A hexagon grid.<\/strong> The notebook draws a 150 km geodesic circle around patient zero and fills it with H3 hexagons at resolution 7, each about 5.2 km\u00b2. That is roughly 14,000 cells. Hexagons are handy here because every neighbor is about the same distance away, which matters when things move.<\/p>\n

2. People from census tracts.<\/strong> Population comes from Esri’s USA Census Tracts layer<\/a>, read straight from the feature service. Tracts and hexagons don’t line up, so each hexagon gets a share of every tract it overlaps, in proportion to the overlapping area. Tracts are clipped to the coastline, so the Pacific stays empty (zombies can still walk through it).<\/p>\n

3. Vulnerability from the CDC SVI.<\/strong> The CDC\/ATSDR Social Vulnerability Index<\/a> ranks every tract on factors such as crowding, age, disability, and vehicle access. I turned a subset into two indices that change the odds in each cell:<\/p>\n