Biomechanical Gait and Accelerometer Analysis to Anticipate Street Crossings
Analyzing smartwatch wrist swings and smartphone IMU acceleration spikes to predict that a standing pedestrian is about to step off the sidewalk 1.5 seconds early.
Neural Gait State Classifiers
Classifying pedestrian kinetic states (standing, texting, running) generates dynamic collision safety envelopes around distracted smartphone users.
- Predicts crossing intent up to 1,500 milliseconds before physical step.
- Reduces pedestrian collisions in vehicle blind spots by over 92%.
- Autonomous pre-charging of hydraulic brake lines.
- Zero false emergency stops on pedestrians remaining safely on the sidewalk.
Explore Domain Portfolio