• ASSET LIST
  • LOCATION
  • VALIDATION
  • GAP ANALYSIS

How to audit street infrastructure with your AI

Connect Mapillary to your agent to turn street-level imagery into structured data. Once connected, your AI can scan specific areas to find and count road assets.

Ask AI about this page

Short answer

How can I use AI to audit street infrastructure?

Your AI uses Mapillary imagery to find specific road features like signs or markings in a given area. It identifies these objects and provides their locations and counts. You'll get a structured list of findings for your audit.

Where every issue lands.

Audit outcomes

The AI organizes its findings into these specific categories.

ASSET LIST

Inventory of features

The AI pulls a list of every detected sign or marking. You get a clean tally of what exists in the area.

LOCATION

Geographic coordinates

Each detected object is tied to its specific place on the map. This lets you pinpoint exactly where a sign is missing or damaged.

VALIDATION

Human review needed

The AI flags detections that need a second look. You decide if the machine's identification of a specific sign type is correct.

GAP ANALYSIS

Missing feature reports

By comparing detections against expected map features, the AI identifies where infrastructure is absent.

The workflow

What your AI does when the request arrives.

The AI moves from broad geographic searches to specific object identification.

  1. Find imagery

    The AI looks for street-level photos within your target area.

    search_images

  2. Identify objects

    The AI scans the images to find specific road features like signs or markings.

    get_image_detections

  3. Extract data

    The AI pulls the specific values or counts for the objects it found.

    get_detection_value

  4. Cross-reference

    The AI checks the detected objects against existing map data to find discrepancies.

    get_map_features

Try it

Copy these to start.

Use these prompts to begin your audit.

Starting points

Swap out the bracketed text for your specific city, street, or feature type.

Mapillary Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_B442verPH186qvVTZbo63Sx80S9hDy2EioTaFfgM/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Mapillary capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "mapillary-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_B442verPH186qvVTZbo63Sx80S9hDy2EioTaFfgM/mcp"
    }
  }
}
Copy into chat04

  • Search for images in downtown Chicago to find stop signs.

  • Find all traffic light detections in the area around Broadway and 42nd Street.

  • Compare the detected road markings in this area to the existing map features.

  • Count how many speed limit signs are visible in these images from San Francisco.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Mapillary once, then ask.

One link connects the MCP to your agent. Your credentials stay encrypted and the setup is handled on our side.

Connect Mapillary to your AI

FAQ

How this task behaves.

  • 01

    Can the AI change the data in Mapillary?

    No. The AI can only read imagery and detection data. It cannot edit or delete anything in your Mapillary account.

  • 02

    Does the AI automatically fix broken signs?

    No. The AI identifies where signs are missing or damaged, but the actual repair work remains a human task.

  • 03

    Can I audit any location in the world?

    The AI can only audit areas where Mapillary has existing street-level imagery.

  • 04

    How accurate are the detections?

    The AI relies on Mapillary's computer vision. You should always review the results for critical infrastructure decisions.

  • 05

    Can the AI create new map features?

    No. The AI can only retrieve existing features or detect new ones in images.

  • More questions about Mapillary? See everything the Mapillary Connector can do

Connect Mapillary to Claude, Cursor, ChatGPT & more