University of Florida (UF) Testbed Initiative – Alternative Transportation Safety Systems

Agarwal, Nithin; Rahmani, Roozbeh; Masciocchi, Jasper · 2020 · ROSA P / Florida Department of Transportation

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 evaluates the safety implications and operational effectiveness of the Mobileye Shield+ Advanced Driver Assistance System (ADAS) on public transit buses. Motivated by rising transit fatalities in Florida and the high liability costs associated with collisions, the research aimed to determine if vision-based ADAS could reduce conflicts between buses, pedestrians, and other vehicles. The study also sought to assess driver acceptance and develop a benefit-cost analysis tool to help transit agencies justify the financial investment in such technology. The research was conducted on 10 Regional Transit System (RTS) buses operating on University of Florida campus circulator routes in Gainesville, Florida. The Mobileye Shield+ system, equipped with camera sensors and driver alert displays, was installed in early 2019. Data collection occurred over approximately one year, divided into two phases: a two-month "stealth mode" where alerts were disabled but data recorded, and a subsequent "open mode" where drivers received visual and audible alerts for events such as pedestrian detection, forward collision warnings, and aggressive braking. Due to equipment malfunctions, data from three buses were excluded, leaving seven buses for analysis. Researchers utilized the Ituran telematics interface and R programming for data cleaning, aggregate analysis, and route-based evaluation. Additionally, five focus group sessions were conducted with bus drivers to gather qualitative feedback, and a macro-enabled Excel tool was developed to perform benefit-cost analyses using surrogate safety measures and historical crash data. The results indicated a significant reduction in traffic conflicts following the activation of the ADAS. In the aggregate before-after analysis, pedestrian-related alerts decreased by 13.3% to 19.6%, while pedestrian collision warnings dropped by 33.4% to 38.6%. Vehicular alerts, including forward collision and headway warnings, decreased by 12.6% to 48.3%, and aggressive braking incidents fell by 29.2% to 47.6%. The average reduction across all warning types was 34.17%, yielding a conflict modification factor of 65.83%. Route-based analysis confirmed these trends, with most routes showing reductions in alerts, though some specific routes experienced slight increases or statistically insignificant changes due to low sample sizes. Hotspot analysis identified 14 high-risk locations characterized by specific infrastructure features like curves, intersections, and pedestrian signal timing. Qualitative feedback from drivers revealed that while they generally preferred the system and found it useful for safety, particularly in congested campus areas, they expressed concerns regarding false positive alerts, sensitivity issues, and a lack of initial training. The study concludes that ADAS implementation effectively reduces transit-related conflicts and improves driver behavior. The developed benefit-cost analysis tool demonstrated that the system offers a positive return on investment for agencies with high historical crash rates, large fleets, and high vehicle revenue miles. The findings suggest that while ADAS is a viable safety enhancement, successful deployment requires addressing driver training needs and managing system sensitivity to minimize false alerts. This research provides transit agencies with empirical evidence and practical tools to evaluate the safety and economic benefits of adopting advanced driver assistance technologies.

Key finding

Mobileye Shield+ on Gainesville RTS buses was associated with roughly 26–34% reductions in recorded transit–pedestrian/cyclist conflicts in before-after analyses, with most focus-group drivers reporting the system useful despite concerns about false alarms and night-time limitations.

Methodology

mixed_methods

Sample size: 10 RTS buses equipped (7 analyzed after 3 units excluded); driver focus groups (~16 drivers queried on automated braking)

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).

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.

• adas effectiveness
• incidence prevalence
• naturalistic crash near crash
• telematics crash prediction

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).

• Applied Guidance: countermeasure evaluation
• Empirical Findings: crash risk outcomes
• Methodological Resource: dataset resource