Hazard-Based Model of Activity Generation Using Vehicle Trajectory Data
DOI: 10.1016/j.procs.2020.03.158
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses the challenge of developing predictive activity generation models for transportation simulation, which traditionally suffer from a lack of longitudinal data due to high survey costs and retention difficulties. While ubiquitous data sources like GPS traces offer longitudinal mobility patterns, they lack the demographic and contextual information necessary for causal modeling. The study aims to bridge this gap by using vehicle trajectory data to develop a hazard-based model of activity generation, enabling the integration of such data into agent-based simulation frameworks. The methodology combines five distinct data sources: vehicle trajectory data (VTD) from 498 GPS-enabled vehicles in Ann Arbor, Michigan; demographic data from the American Community Survey (ACS); land use data; accessibility metrics from the Environmental Protection Agency’s Smart Location Database; and a Household Travel Survey (HTS) from the Chicago Metropolitan Agency of Planning. To infer activity purposes from anonymous VTD, the authors first identified home locations using heuristic algorithms based on frequent visitation and validated them against land use maps. For non-home activities, they developed a multinomial logit model using the Chicago HTS data, utilizing trip characteristics such as distance from home, departure time, and duration to classify 13 activity purposes. This model was then applied to the VTD to label activity purposes. Finally, a hazard-based model was estimated using the inferred longitudinal data to analyze inter-activity durations for nine discretionary activity types. The results demonstrate that the inferred activity purposes and durations closely matched those from prompted recall surveys. The hazard-based model, estimated on 34,442 observations, revealed that all nine activity types exhibited a decreasing failure rate over time. Specific demographic and accessibility factors significantly influenced activity generation. For instance, low-income households showed a higher hazard for personal business activities, while higher transit accessibility increased the likelihood of personal business and shopping activities. Shopping activity was less likely among middle-aged and elderly individuals compared to younger groups, whereas higher income levels correlated with increased shopping probability. Recreation hazards were higher for younger individuals and those with middle incomes, and were positively impacted by transit accessibility. Visiting activities were influenced by gender, age, income, race, and household structure, with medium-income households showing higher hazards than both low- and high-income groups. The significance of this work lies in its demonstration that vehicle trajectory data, when augmented with external demographic and land-use datasets, can effectively support the development of causal activity generation models. By overcoming the limitations of cross-sectional data, this approach allows for the capture of intra-individual variability and the dynamics of activity engagement. This enables more accurate long-term forecasting in agent-based transportation simulations, advancing the utility of ubiquitous data sources in travel demand modeling.
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 | Crossref | — | — | 1 | 2026-06-20 |
| archive | success | openalex | — | — | 5 | 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 |
| enrich | success | openalex | — | — | 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.
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: dataset resource
- Theoretical Contribution: computational model