SAFE-D: A Spatiotemporal Detection Framework for Abnormal Driving Among Parkinson's Disease-like Drivers
DOI: 10.48550/arXiv.2510.17517
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 introduces SAFE-D, a spatiotemporal detection framework designed to identify abnormal driving behaviors caused by Parkinson’s disease (PD). While existing research focuses on temporary functional impairments like drowsiness or distraction, there is a significant gap in detecting anomalies stemming from chronic pathological conditions. Given that approximately 60% of individuals with PD actively drive, the authors aim to elucidate how PD-induced motor dysfunction manifests in vehicular control and to establish a detection mechanism for these pathology-specific behaviors. The methodology begins by analyzing PD symptomatology, identifying tremors, rigidity, and bradykinesia as the primary motor symptoms impacting driving. The authors establish causal links between these symptoms and specific driving impairments: tremors cause continuous fluctuations in steering control during straight-line driving, while rigidity and bradykinesia lead to sudden, unstable variations in coordinated control during non-straight maneuvers like turns. To capture these signatures, the framework integrates data from three in-vehicle sensors: steering wheel torque, accelerator pedal pressure, and brake pedal pressure. The system employs an attention-based neural network that adaptively prioritizes spatiotemporal features. Data preprocessing includes normalization, frequency standardization to 30 Hz, and segmentation into 4-second windows. The model uses a ResNet module for feature extraction and a channel-aware multi-head attention mechanism to fuse global and local spatiotemporal information, enhancing sensitivity to motor control abnormalities. Experimental validation was conducted using a Logitech G29 driving system integrated with the CARLA simulator. The dataset comprised driving data from three representative road maps—urban, rural, and mixed environments—to emulate real-world conditions. The framework was evaluated on its ability to distinguish between normal driving patterns and those affected by PD. Results demonstrate that SAFE-D achieves an average accuracy of 96.8% in detecting PD-related abnormal driving behaviors. The model successfully identified the distinct behavioral signatures associated with PD motor deficits, such as continuous steering fluctuations and abrupt pedal variations, without requiring auxiliary hardware. The significance of this work lies in bridging the gap between clinical PD diagnosis and dynamic driving safety monitoring. By leveraging standard vehicle interfaces, SAFE-D offers a plug-and-play solution for continuous, non-invasive monitoring of neurological risks during routine commutes. This approach provides complementary behavioral evidence to enhance diagnostic accuracy and offers actionable insights for personalized healthcare. The study underscores the potential of integrating clinical pathophysiology with vehicular telemetry to improve public safety for drivers with neurodegenerative disorders.
Key finding
The SAFE-D framework achieves 96.8% average accuracy in distinguishing between normal and Parkinson's disease-affected driving patterns using spatiotemporal feature fusion from vehicle control sensors.
Methodology
simulation_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. Discovered via openalex_abstract on 2026-05-08.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | — | — | — | 1 | 2026-05-07 |
| archive | success | canonical_url | — | — | 1 | 2026-06-06 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | openalex | — | — | 2 | 2026-05-08 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 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).
- Theoretical Contribution: computational model