# A Behavioral Ground Truth for Exteroceptive Sensors: Geometric Constraints and Stochastic Duration in Parking Maneuvers

## TL;DR

This study creates a dataset of parking maneuvers to improve autonomous vehicle sensors by capturing human behavior patterns and timing.

## Contribution

The paper introduces a stochastic human baseline dataset for parking maneuvers, revealing geometric and temporal patterns for sensor calibration.

## Key findings

- 45° angled parking is the most efficient, with an average maneuver time of 7.54 seconds.
- Narrow aisles cause significant delays in entry, with 85th percentile times exceeding 50 seconds.
- Exit maneuvers take about 54% less time than entry, highlighting functional asymmetry in parking behavior.

## Abstract

What are the main findings?
Aisle width in parking maneuvers generates distinct non-linear kinematic signatures, offering essential behavioral patterns for onboard sensor training.Furthermore, the temporal asymmetry observed between entering and exiting the parking lot defines dynamic tolerances, providing key parameters for obstacle tracking by autonomous sensors.

Aisle width in parking maneuvers generates distinct non-linear kinematic signatures, offering essential behavioral patterns for onboard sensor training.

Furthermore, the temporal asymmetry observed between entering and exiting the parking lot defines dynamic tolerances, providing key parameters for obstacle tracking by autonomous sensors.

What are the implications of the main findings?
The empirical ground truth dataset on parking dynamics enables the fine-tuning of exteroceptive sensors, preventing human hesitations from being classified as static obstacles.Integrating these metrics optimizes predictive logic, improving the reliability of sensory detection for autonomous parking systems in shared urban spaces.

The empirical ground truth dataset on parking dynamics enables the fine-tuning of exteroceptive sensors, preventing human hesitations from being classified as static obstacles.

Integrating these metrics optimizes predictive logic, improving the reliability of sensory detection for autonomous parking systems in shared urban spaces.

The deterministic simplification of parking maneuvers in traditional traffic models presents a critical challenge for the safe integration of Autonomous Vehicles (AVs). This study establishes a stochastic human baseline to provide a naturalistic ground truth dataset essential for calibrating perception and prediction sensors in mixed traffic scenarios. Through the analysis of 1038 maneuvers observed in a university shared space in Catania, Generalized Linear Models and Kaplan–Meier estimators were applied to quantify the impact of geometric constraints on 0°, 45°, and 90° configurations. Results identify 45° angled parking as the Pareto-optimal solution regarding stability and speed, achieving an average maneuver time of 7.54 s. Furthermore, a vertical parking paradox emerges: in the presence of narrow aisles, entry times increase drastically, generating bottlenecks with an 85th percentile exceeding 50 s. Finally, a structural functional asymmetry reveals that exit maneuvers require approximately 54% of the time needed for entry. These findings provide empirical metrics essential for validating human behavior models and fine-tuning decision-making and timeout logic in autonomous driving systems.