What AI should do here
For each walk, AI drafts captions, extracts visible public-realm tags, compares repeat photos, and flags possible issues for human review.
One-Street AI Atlas
A monthly, privacy-first field notebook for the 1300 block of Cumberland Avenue near Colvin Street: repeat photos, public records, and AI-assisted observations focused on what changes at street scale.
The walk
Each monthly pass uses the same photo stops so small changes become visible over time.
Photo stop 1
Capture the approach into the 1300 block, sidewalk condition, lighting, tree canopy, drainage, and signs.
For each walk, AI drafts captions, extracts visible public-realm tags, compares repeat photos, and flags possible issues for human review.
Blur faces and license plates before publishing. Avoid house-by-house judgment. Treat public records as context, not as proof of current conditions.
Take the same three views, add weather and time of day, note changes, check nearby records, then publish a short narrative update in data/atlas.json.
Photo log
Initial phone photos document the repeat-view baseline. Photo files still need to be copied into the project folder before the published page can show the original images.
Colvin corner
South-to-north view looking out over Colvin, useful for the Cumberland/Colvin intersection context.
Cumberland view
North-to-south view looking up Cumberland from near Colvin.
Cumberland view
South-to-north view down Cumberland showing the everyday neighborhood street context.
Pavement condition
Close-up of pavement cracking on Cumberland, including edge cracking and loose organic debris on the road surface.
Issues nearby
Filters separate public-realm concerns from property-context records so the story stays careful and useful.
SYRCityline / 311
Look for recent requests near the block: potholes, sidewalk concerns, street lights, illegal dumping, snow, drainage, or traffic-control issues.
Open SYRCityline sourceCode Violations V2
Query nearby records by address or buffer. Show status, date, and category only when it adds context to a visible block-level pattern.
Open feature layerHuman-reviewed AI
AI may flag cracked pavement, blocked sightlines, poor lighting, missing curb ramps, water pooling, or repeated trash conditions for manual confirmation.
Living narrative
The project becomes stronger as the same observations repeat across weather, seasons, and city-service cycles.
May 2026
Initial Cumberland/Colvin field photos were reviewed: two directional street views, one long block baseline, and one pavement-crack detail. Original image files still need to be copied into the project folder for publication.
Next monthly walk
Copy original image files into the project assets folder, blur any faces or plates if needed, then replace the photo placeholders with the real images and repeat the same views next month.
Sources
Every public-data claim should keep a source link and a last-checked date.
City portal for public datasets, maps, and visualizations.
View sourceArcGIS item for service requests generated through QAlert. Use for nearby issue checks where available.
View sourceCity code-enforcement records from 2017-present, described by the source as daily updated.
View source