evidoria

← Back to browse

Good practice Imported

City Detect — Dallas Puts AI Cameras on Trash Trucks to Flag Illegal Dumping and Blight

United States of America · Dallas · See the United States of America profile

Evidence: Descriptive / self-reported Top 32% 67/100 · Ask Evidence Copilot about this practice

Dallas mounted 100 AI cameras on 50 sanitation trucks under a $2.556 million, three-year contract with City Detect Inc. to scan neighbourhoods for illegal dumping and blight; it flagged over 21,000 properties and issued about 1,800 courtesy notices in its first months.

$2.556 million over 3 years (~$852,000/year)
Contract value (approved Dec 2025)
100 cameras on 50 trucks
Cameras and trucks deployed (as of 2026)
>21,000
Properties flagged (by September 2026)
~1,800
Courtesy notices issued (by September 2026)
City Detect — Dallas Puts AI Cameras on Trash Trucks to Flag Illegal Dumping and Blight

Details

Maturity
Scaling
Promoter
City of Dallas Sanitation Services / City Detect Inc.
Period
2025-2026
Keywords
code enforcement, sanitation, computer vision, municipal government

Context

Dallas sought a way to systematically identify illegal dumping, graffiti, overgrown vegetation and deteriorating structures across the city using its existing sanitation-truck routes.

Objectives

Use AI-linked cameras mounted on collection trucks to flag blight and dumping for code-compliance follow-up, with human staff reviewing every detection before any notice or citation.

Activities

In December 2025, Dallas City Council approved a three-year, $2.556 million contract (about $852,000 a year) with City Detect Inc. Two cameras were fitted to each of 50 brush/bulky-waste collection trucks, photographing properties on routine collection runs roughly every 30 days. Computer-vision software flags illegal dumping, graffiti, overgrown vegetation and deteriorating structures, assigning a 'blight score' from 1 to 4; faces and license plates are automatically blurred, and Dallas code-compliance staff review every AI-flagged detection before any notice or citation is issued.

Results

By September 2026, reporting put the number of flagged properties at more than 21,000, with about 1,800 resulting in courtesy notices rather than formal citations. The same vendor has run a similar system for Columbia, South Carolina since 2024, with comparable systems in Tacoma, Washington and San Jose, California.

Conclusions

The programme has generated substantial detection volume in its first months, but independent reporting has not yet evaluated whether flagged properties translate into a measurable drop in dumping, blight or code-violation backlogs, and at least one council member has already proposed cutting its funding.

Implementation

Indicative cost
Medium (€50k–€500k) — $2.556 million three-year vendor contract (~$852,000/year) for 100 cameras across 50 trucks plus computer-vision analytics and review workflow.
Time to results
Medium (1–3 years) — Contract approved December 2025; three-year term; over 21,000 properties flagged within roughly the first nine months of operation.
Staffing & skills
City of Dallas Sanitation Services, Dallas code-compliance review staff (review every AI-flagged detection), vendor City Detect Inc.

Conditions for success

  • Human staff review every AI-flagged detection before any notice or citation is issued -- the AI does not act on its own
  • Automatic blurring of faces and license plates to limit privacy exposure
  • Reuse of existing sanitation-truck routes rather than dedicated patrol vehicles

Common failure modes

  • At least one Dallas council member proposed cutting the programme's funding within months of launch
  • No independent evaluation yet of whether flagged properties translate into reduced dumping, blight or code-violation backlogs
  • Reporting has raised questions about which neighbourhoods receive the most scrutiny

Where it fits

Governance type
municipal government
Scale
citywide, 50 trucks
Income level
high-income

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful