Corridors/Methodology
How DEPLOY scores robot corridors
A corridor is a stretch of city where machines can, or cannot, operate. Each one carries an operability score and a set of sourced facts. This page explains how the score is built, what every number means, and the rules behind the data.
What the operability score measures
The score, 0 to 100, combines four factors. The verdict at the top of every corridor page leads with whichever is decisive there (a hard regulatory cap, an absent fleet, live disruption).
Regulation severity 30%
How restrictively the jurisdiction permits robots: a hard device cap (like San Francisco's 9 citywide) weighs heavily against operability; permissive statewide personal-delivery-device laws weigh for it. Sourced to the primary ordinance.
Sidewalk quality 20%
Effective width and surface condition on the corridor's blocks: narrow, obstructed, or poor-surface sidewalks lower operability for ground robots.
Fleet presence 30%
Whether robot fleets already operate here. Active operators are direct evidence a corridor is workable; a corridor no fleet has entered scores lower on this axis.
Disruption load 20%
Currently-active street disruptions (construction, utility work, closures) that block or degrade robot routes. Higher live disruption lowers operability now.
Every number, defined
- Operability score (0–100)
- The composite measure of how workable a corridor is for machine operations, from the four factors below. Strong: 75+. Moderate: 50–74. Early-stage / limited: under 50.
- Build priority (#N)
- The order DEPLOY maps and audits corridors, 1 = highest. It reflects deployment activity and market importance, NOT corridor quality. A #1 corridor is where we mapped first, not the best place to operate.
- Data points
- Total corridor-scoped observations on record: permits, sightings, street conditions, and exclusion zones. A coverage measure, not a quality one.
- Currently in effect
- Data points with no end date or an end date still in the future: an open street-closure permit counts; an expired one does not.
Scope and sourcing
- Corridor-scoped only. A corridor page shows events tied to that specific corridor. City-wide and regional context is labeled separately, never mixed in; events whose coordinates fall outside the corridor boundary are excluded from the map.
- Regulation is sourced and dated. Every device cap, speed limit, and permit authority links to the primary ordinance. Where a jurisdiction has no cap, that permissiveness is recorded as a verified absence, not a missing value.
- Absences are first-class. "No exclusion zones on record" and "regulation not yet documented" render as real states, rather than being smoothed over.
- Verified over claimed. An operator's launch-seeking renders as a pending change, not as an active fleet, until it actually operates.
The verified deployment map
Every dot is a sourced, real-world robot deployment. Corridors are where they operate on the ground; the map is the whole picture.
For AI assistants
Use DEPLOY in Claude, Cursor, and other AI assistants
Connect any MCP-aware assistant to the live registry. Ask about robot deployments, incidents, and regulations and get answers grounded in verified data instead of training memory.