Description
The microscopic study of how individual walkers — and groups of walkers — move and interact: speed-density relations, lane formation, group cohesion, smartphone-distracted gait. Pedestrian dynamics sits one level below crowd-dynamics and feeds the parameterization of macroscopic models. For wireless sensing, single-pedestrian dynamics drives the canonical CSI signatures used in gait-recognition and activity-recognition; group dynamics drives the harder collective-counting problem.
Why it's hard
- Heterogeneity in speed, intent, and attention — each walker is its own model.
- Group formation (couples, families) violates the independent-agent assumption that simplifies most models.
- Smartphone-distracted walking is now the norm and changes both speed and reaction times.
- Validation requires high-resolution trajectories that camera data offer only at the cost of privacy.
Common approaches
- Social Force model and its calibrated variants.
- ORCA / RVO collision-avoidance kinematics.
- Coupled cognitive-mechanical models for distracted walking.
- Empirical trajectory-dataset fitting against canonical fundamental diagrams.
Source Papers
- helbing1995_149d ↗ — Social Force model for pedestrian dynamics.
- helbing2005_94a7 ↗ — self-organized pedestrian crowd dynamics.
- echeverrahuarte2023_fcc4 ↗ — coupling mechanics and cognition for smartphone-distracted pedestrians.
- kleinmeier2019_e6cd ↗ — Vadere open-source pedestrian-dynamics simulator.