AI & data standards

Better infrastructure with privacy in mind.

CYVL exists to make infrastructure better — not to track people. We collect data on roads, sidewalks, and assets, and we deliberately build the technology so faces, license plates, and personal identities are stripped out before anything reaches a customer.

Serving · public agenciesPrivacy-first
Residents walking into a brick city hall with the American flag flying out front

Trust & data privacy

We map infrastructure, not people.

Big technology companies map streets to learn addresses and track where people go. We don't. All we care about is making infrastructure better — that's core to our mission, and we build the technology so your residents' privacy is protected by default.

De-identified by default

Roads, not faces

Faces and license plates are blurred before imagery is ever delivered. What a city receives is condition data and geospatial layers — not photographs of residents.

Scoped models

AI within limits

Our models see pavement, signs, and assets — not people. We don't extract biometric identifiers, and we don't cross-reference license plates against owner records.

Category-defining

A published standard

CYVL publishes the Municipal Infrastructure Data Privacy Standard — a category-defining bar for how municipal road data should be handled, adoptable directly into procurement.

The standard

The Municipal Infrastructure Data Privacy Standard.

A category-defining standard for vendors collecting roadway imagery on behalf of cities, transportation agencies, and public-sector clients — adoptable directly into procurement requirements.

The five commitments

01

Faces and license plates are automatically blurred in all imagery before delivery.

De-identification runs before any imagery is released to the customer-facing platform or included in any deliverable.

02

License plates are not cross-referenced against DMV or vehicle registration records.

No integration with motor vehicle databases. No lookup of plate numbers against owner identities.

03

No biometric identifier templates are extracted from any captured imagery.

No facial recognition templates, gait analysis, or other biometric identifiers are derived from collected imagery.

04

No deliverable contains personally identifiable information.

Customer-facing outputs consist only of condition data, asset inventories, and geospatial layers — numerical and geospatial outputs.

05

All collection complies with the federal Driver's Privacy Protection Act.

DPPA compliance is enforced architecturally, not just by policy.

Why this matters

Municipal infrastructure data collection is becoming more capable every year. Vehicles equipped with cameras, LiDAR, and AI can capture and analyze every public mile in a city. The technology is valuable. The privacy implications, if not addressed deliberately at the architectural level, are not.

Privacy and data utility are not opposed. Modern computer vision can extract everything a city needs to know about pavement, signs, and assets from imagery that has been stripped of identifying features. The privacy is a consequence of building the technology correctly, not a cost imposed on top of it.

This standard sets the bar for what every municipal road data vendor should be required to meet. CYVL meets all five commitments today. Cities adopting this language into procurement specifications can require the same of any vendor they engage.

American-made

Built in America, end to end.

Our entire supply chain is American. Every sensor, every line of on-device software, every server, and every engineer is based in the United States — a distinction that matters for the agencies we serve.

The American supply chain

01

Sensors built, assembled, and manufactured in the United States.

Every CYVL sensor rig is produced domestically — not sourced or assembled overseas.

02

Defense-grade LiDAR and cameras, approved for U.S. government use.

The LiDAR units and cameras we deploy are approved by the U.S. Department of Defense — no foreign or restricted hardware in the field.

03

On-device software written in the United States.

The software running on every sensor is developed in-house by our own team, not licensed from foreign vendors.

04

Data stored on U.S.-based cloud infrastructure.

Captures are uploaded to and stored on cloud infrastructure located in the United States — your city's data never leaves the country.

05

Engineering and development team based in Boston, Massachusetts.

Our entire team of engineers and developers works from Boston — American-built, by Americans.

Why this matters

Many infrastructure-data vendors are headquartered overseas and rely on foreign hardware — in some cases Chinese. That introduces questions about data sovereignty, supply-chain security, and long-term support. CYVL was built to answer them: from the sensor to the server to the software, everything is American.