SteelCollarRobotics and Aeronautics
Ground JoeAir AmeliaOrbit KatherineGovernments Abe
Home / The casebook / Public agencies

Evolv weapons screening pilot in the New York City subway

Municipal police department screening a transit system · Weapons screening · Trade press

An independent editorial outlet wrote its own account of this deployment. Numbers still appear exactly as published, and still need verification against your own job.

The job

The New York Police Department ran a 30-day pilot of Evolv walkthrough weapons scanners at subway station entrances starting July 27, 2024, rotating the units across 20 stations. Officers screened riders passing through the portals on a voluntary basis, and the department released the outcome figures roughly two months after the pilot ended, following repeated press requests.

The agency

20 subway stations hosted the rotating pilot (reported-by-trade-press); systemwide scale: not published in the cited accounts

The equipment

Evolv walkthrough weapons detection scanners (reported-by-trade-press)

The published numbers

2,749 total scans over 30 days at 20 stations (reported-by-trade-press); 0 guns found (reported-by-trade-press); 12 knives found (reported-by-trade-press); 118 false positives (reported-by-trade-press); the NYPD did not pay for the pilot and had not contracted with Evolv at the time of reporting (reported-by-trade-press); pilot cost to the city: not published

Where a source did not publish a figure, this entry says so rather than guessing.

What we would check before believing it applies to you

This entry is included as a caution. The pilot's own numbers, zero guns and 118 false positives across 2,749 scans, led the Legal Aid Society to call it objectively a failure, and the data came out only after sustained press requests, which is itself a transparency lesson. Screening was voluntary and riders could enter at another turnstile, so the counts cannot establish much either way in an open transit system. Evolv as a company has faced federal regulatory scrutiny and litigation over claims about what the technology can actually do, which any buyer must weigh independently of a vendor demonstration. An agency considering walkthrough screening should define its success metric before the pilot, commit to publishing results on a schedule, and model false-positive handling at realistic throughput, because secondary screening is where the staffing cost and the civil liberties risk both concentrate.

Sources

Is your job like this one?

Describe what your agency is weighing. Abe will give you an honest preliminary assessment, and if a robot or a drone is the wrong answer, he will say so.