ANALYZE THE SPATIAL DISTRIBUTION OF DELIVERY MOTORCYCLE CRASHES AND IDENTIFY THE RELATED FACTORS
DOI: 10.5194/isprs-archives-XLIII-B4-2022-163-2022
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 investigates the spatial distribution of delivery motorcycle crashes in Taipei, Taiwan, and identifies associated environmental factors. Motivated by the surge in online food delivery services during the COVID-19 pandemic and the resulting increase in traffic accidents, the research addresses a gap in existing literature that often overlooks the distinct driving behaviors of delivery riders or fails to account for spatial dependencies in crash data. The authors aim to determine how land use and points of interest (POIs) influence crash frequency, specifically comparing the performance of traditional global models against local spatial models. The methodology utilizes data from 2,314 delivery motorcycle crashes recorded in Taipei during 2020. The study area is divided into 456 villages, with explanatory variables including POI counts for restaurants, supermarkets, shopping malls, schools, hotels, and bus stops. Two statistical models were employed: a Generalized Linear Model (GLM) using negative binomial regression to handle overdispersion, and a Geographically Weighted Negative Binomial Regression (GWNBR) model to capture spatial heterogeneity. The GWNBR allows coefficients to vary spatially, addressing the non-stationary nature of crash data. Model performance was evaluated using the Akaike Information Criterion (AIC), log-likelihood, and Root Mean Square Error (RMSE), while spatial autocorrelation in residuals was assessed using Moran’s I statistics. The results demonstrate that the GWNBR model significantly outperformed the GLM across all performance metrics, including lower AIC and RMSE values and higher log-likelihood. Furthermore, the GWNBR residuals showed insignificant spatial dependency, indicating that the model effectively accounted for spatial heterogeneity. In terms of crash factors, the number of restaurants was found to have a significant positive association with crashes on both weekdays and weekends. Other commercial POIs, such as supermarkets, shopping malls, and bus stops, showed significant associations only during weekends. Spatial analysis revealed that the impact of these factors intensified in suburban and rural areas compared to urban centers. For instance, the coefficient for restaurants increased from urban to rural areas, suggesting that adding restaurants in less dense areas poses a higher relative crash risk, potentially due to weaker traffic policies and infrastructure. The study concludes that incorporating spatial heterogeneity and overdispersion into crash modeling provides superior predictive accuracy compared to global models. The findings highlight that commercial areas and bus stations are critical risk factors, with effects exacerbated in rural and suburban zones. These insights are significant for traffic safety planning, suggesting that authorities should prioritize safety interventions in non-urban areas where delivery activity intersects with commercial infrastructure. The research underscores the need for tailored traffic policies that account for the specific spatial dynamics of delivery motorcycle operations, particularly as the popularity of food delivery services continues to grow.
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 | DOAJ | — | — | 1 | 2026-06-19 |
| archive | success | unpaywall | — | — | 1 | 2026-06-25 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-19 |
| chunk | success | chunk | — | — | 1 | 2026-06-19 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-19 |
| promote | success | — | — | — | 1 | 2026-06-19 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-19 |
| 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.
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).
- Empirical Findings: crash risk outcomes