Pilot Application of Biometric-Based Vehicle Occupancy Detection on Managed Lanes for Congestion Reduction
archive: archived pipeline: cataloged verified
Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)
Summary
This study evaluated the accuracy and reliability of RideFlag, a biometric-based smartphone application designed to detect vehicle occupancy for High Occupancy Vehicle (HOV) and High Occupancy Toll (HOT) lane management. The research was motivated by the inefficiency, safety risks, and high violation rates associated with current enforcement methods, such as manual law enforcement stops and honor-system transponders, which often result in significant HOV misdeclaration. The goal was to determine if an in-vehicle, app-based solution could accurately distinguish between single-occupant vehicles and carpools to support congestion reduction and incentivize carpooling. The pilot study, conducted from July 2022 to February 2023, involved 29 volunteer carpoolers from South Florida, the Tampa Bay area, and Utah who logged 837 trips using the app. Participants used the application to capture biometric facial data at the start and end of trips within geofenced highway segments. The app utilized artificial intelligence to assess "realness" (distinguishing human faces from replicas) and "similarity" (matching facial geometry between trip start and end). Independent evaluators validated the app’s determinations by comparing them against digital photos captured simultaneously. The study also included staged tests where evaluators attempted to deceive the system using single-occupant vehicles. The results demonstrated high accuracy in occupancy detection. Of the 837 logged trips, 648 were identified as True Positives (correctly validated carpools), 121 as True Negatives, 1 as a False Positive, and 13 as False Negatives. The app achieved a Positive Predictive Value of 0.9985 and a Negative Predictive Value of 0.9030. Sensitivity was 0.9803, and specificity was 0.9918. Most False Negatives were attributed to an overly strict similarity threshold, which is configurable. In staged tests, evaluators failed to trick the app into validating a single-occupant vehicle as a carpool after more than 30 attempts. The app’s performance improved throughout the pilot, and user feedback indicated acceptable usability, though users noted the need to remember to capture initial snapshots and wait for verification signals. The findings suggest that biometric-based vehicle occupancy detection is a viable alternative to traditional enforcement methods, offering high accuracy and reduced reliance on manual verification. The study concludes that the technology requires a policy decision regarding the balance between validation speed and accuracy, specifically concerning similarity thresholds. A soft launch period is recommended to calibrate these thresholds and manage the trade-off between False Positives and False Negatives. Future research should focus on integrating the app with back-office systems for automated reward delivery, testing accuracy for vehicles with three or more occupants, and conducting social marketing studies to identify barriers to adoption. The technology has already informed the development of subsequent app versions for broader pilot programs.
Key finding
The biometric vehicle occupancy detection app demonstrated high accuracy in validating carpool trips, achieving a positive predictive value of 0.9803 and a negative predictive value of 0.9030 across 837 logged trips.
Methodology
field_study
Sample size: 29
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | rosap | — | — | 2 | 2026-05-23 |
| archive | success | — | — | — | 1 | 2026-05-23 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-23 |
| promote | success | — | — | — | 1 | 2026-05-23 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 3 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 24 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Methodological Resource: validation psychometrics