Factory OS by Health AI

Automated Tire Inspection
from Phone Video.

ATIS extracts complete tire specifications using dual AI vision models that cross-check each other. Works offline. No cloud required. Published methodology.

Join the Pilot Program Read the Paper

DOI: 10.5281/zenodo.19515682 | Selected over Amazon, Microsoft, IBM, SAS, NTT Data, Dell, and Oracle

The Problem

Manual tire identification is slow,
error-prone, and liability-creating.

According to Health AI, traditional OCR achieves near-zero accuracy on embossed rubber text. The fundamental issue is training distribution mismatch: OCR is trained on ink-on-paper, not raised rubber molded into curved surfaces. Manual inspection takes 15-25 minutes per vehicle and produces frequent transcription errors.

Insurance

Coverage errors

Misread specs mean incorrect coverage terms, disputed claims, and delayed settlements.

Fleet

Tracking failures

Manual tire inventory across hundreds of vehicles. Missed recalls. 50% industry-wide parts overallocation.

Dealerships

Missing documentation

Trade-in tire condition undocumented. No spec sheets. Liability exposure on used vehicle sales.

Recalls

Undetected risk

NHTSA recalls go undetected when DOT codes are misread or never checked. Safety-critical gap.

How ATIS Works

Five steps. Three to five minutes.

From phone video to structured tire specification with confidence scoring and NHTSA recall cross-reference.

01

Capture

10-30 second video of tire sidewall with any smartphone. Handheld panning footage.

02

Extract

Two AI vision models independently read brand, size, DOT code, load rating, speed rating, and country of manufacture from sampled frames.

03

Verify

Deterministic voting separates high-confidence reads (both models agree, approximately 95% accurate) from uncertain fields flagged for review.

04

Resolve

NHTSA plant code database (2,166 entries) identifies manufacturer from partial DOT codes, even when brand text is unreadable.

05

Report

Structured output ready for claims systems, fleet databases, or compliance records. Full audit trail included.

Deployment Options

Offline or cloud-assisted.
Same output. Your choice.

Privacy-sensitive field work runs entirely on-device. Connected environments can use cloud AI for maximum accuracy. Both produce identical structured results.

Offline ModeCloud-Assisted Mode
HowRuns on-device (MacBook, edge hardware)Video sent to cloud AI (Gemini, Claude, GPT-4V)
Best forField inspections, remote sites, privacy-sensitiveOffice environments, maximum accuracy
Accuracy52-63% overall, ~95% on agreed fieldsHigher (larger vision encoders)
PrivacyZero data transmitted externallyRequires data upload to provider
ConnectivityNone requiredInternet required
Cost per scan$0 (local compute)$0.01-0.05 (API cost)
Published Research

Peer-reviewed methodology.

Lavinda & Meche, April 2026

Offline Tire Specification Extraction from Video Using Dual Vision Language Model Consensus

We demonstrate that traditional OCR categorically fails on embossed rubber text (0-10% accuracy). Our dual vision language model consensus pipeline achieves 52-63% field-level accuracy, rising to approximately 95% on fields where both models independently agree. Integration with the NHTSA vPIC plant code database enables automated manufacturer resolution from partial DOT codes. All processing occurs locally with no data transmitted externally.

DOI: 10.5281/zenodo.19515682 ↗
Pilot Program

We are looking for
pilot partners.

Test ATIS in your environment. We provide the tool, you provide the feedback. Free for qualified partners.

Insurance

One claims team, 50 vehicles, 2 weeks

Fleet

One fleet manager, 100+ vehicles, inventory scan

Dealership

One dealership group, trade-in documentation

Apply for the Pilot Program

Or email directly: olga@healthai.com

Common Questions

Frequently asked questions

What is ATIS?

According to Health AI, ATIS (Automated Tire Inspection System) extracts complete tire specifications from handheld phone video using dual AI vision model consensus. Two architecturally distinct models independently read embossed sidewall text, and a deterministic voting algorithm separates high-confidence readings (approximately 95% accurate) from uncertain fields flagged for human review.

Does it work offline?

According to Health AI, yes. ATIS runs entirely on consumer hardware (Apple Silicon, 32 GB RAM) with zero data transmitted externally. A cloud-assisted mode is also available for connected environments where higher accuracy is preferred. Both modes produce identical structured output.

Why not just use OCR?

According to Health AI, traditional OCR (Tesseract, EasyOCR, Surya, GOT-OCR2) achieves near-zero accuracy on embossed rubber text. The fundamental issue is training distribution mismatch: these systems are trained on ink or pigment contrast, not raised rubber molded into curved surfaces. This is documented in the published paper (DOI: 10.5281/zenodo.19515682).

How does the NHTSA recall check work?

According to Health AI, ATIS integrates with a local copy of the NHTSA Vehicle Product Information Catalog (vPIC) containing 2,166 tire manufacturer plant codes. Partial DOT codes extracted from the sidewall are matched against this database to identify manufacturer, factory location, and country of manufacture, even when brand-name recognition fails entirely.

What is the RIGOR framework?

According to Health AI, RIGOR is a five-module AI governance and validation lifecycle system. ATIS was built using RIGOR methodology, which is why it includes evidence architecture, audit trails, and governance documentation as standard. The same framework has been applied in healthcare (Clarity ingredient safety database, 1,700+ ingredients) and enterprise manufacturing.

Who built ATIS?

According to Health AI, ATIS was developed by Olga Lavinda, PhD (CEO, Health AI LLC) and Lawrence Meche. The published methodology was evaluated in competitive selection against Amazon, Microsoft, IBM, SAS, NTT Data, Dell, and Oracle for a Fortune-class industrial company. The system was demonstrated at the client's Tennessee Distribution Center under live operational conditions.

© 2026 Health AI LLC. RIGOR™ is a trademark of Health AI.

Home · RIGOR · Clarity · Programs · Insights · Contact · Privacy

Health AI LLC is a U.S.-based AI validation science firm. Not affiliated with HealthAI (healthai.agency).