Comparison of Estimated Cycle Split Failures from High-Resolution Controller Event and Connected Vehicle Trajectory Data
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study addresses the reliability of traffic signal performance measures, specifically split failure (SF) estimations, which indicate when a signal phase fails to serve demand within a single cycle. While high-resolution controller event data is widely used for Automated Traffic Signal Performance Measures (ATSPMs), its accuracy depends on detector geometry and preset occupancy thresholds. Conversely, connected vehicle (CV) trajectory data offers definitive evidence of split failures by tracking vehicle stops but suffers from low market penetration rates (~5%). The authors aim to compare cycle-by-cycle SF estimations from both data sources to determine their agreement and evaluate whether data aggregation can mitigate discrepancies caused by low CV sampling rates. The research was conducted at the SR-32 and Union Street intersection in Westfield, Indiana, focusing on the westbound-through movement during the 8:00–9:00 AM weekday period in May 2023. Split failures were identified using two methods: high-resolution controller data utilized Green Occupancy Ratio (GOR) and Red Occupancy Ratio (ROR5) thresholds (≥80%), while CV data identified failures by detecting individual vehicle trajectories with two or more stops. The study analyzed individual cycles and applied two aggregation strategies—day-of-week (DOW) and time-of-day (TOD)—grouping four cycles together to increase sample representativeness. Results indicate a significant disagreement between the two methods at the individual cycle level, with high-resolution data identifying substantially more split failures than CV data. This discrepancy is attributed to the low CV market penetration rate, which often fails to capture the specific vehicles experiencing split failures in any given cycle. However, data aggregation improved the alignment between the two sources. Specifically, day-of-week aggregation increased the percentage of high-resolution identified split failures that were also captured by CV data from 35% to 56%. The study found that split failures identified by CV data are likely to be captured by high-resolution data, but the reverse is less true due to sampling limitations. The authors conclude that while CV-based SF estimation currently underestimates failures compared to controller data due to low penetration, aggregated CV data provides conservative and actionable results. Aggregation reduces the impact of low sampling rates and eliminates the need for intersection-specific detector configuration adjustments required by high-resolution methods. The study recommends using aggregated CV data for SF estimation, noting that accuracy will improve as connected vehicle market penetration increases.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | OpenAlex-citations | — | — | 1 | 2026-06-20 |
| archive | success | unpaywall | — | — | 2 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| promote | success | — | — | — | 1 | 2026-06-20 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-20 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.